365Papers
Starting August 22, 2017, I'll read one academic paper every day for 365 days. The topics are intended to span Machine Learning, Neuroscience, Computer Vision, and Computational Medicine, with other topics thrown in liberally for seasoning.
Get in touch with me on Twitter as I tweet #365papers, or follow here. Suggest papers! Read along! Condescend!
This website is hosted by GitHub Pages, and the code is available here.
20 — Coulomb GANs: Provably Optimal Nash Equilibria via Potential Fields
Modeling a GAN target distribution as an electric field removes all but one single Nash equilibrium, improving training efficacy and output quality. [ Read More ]
19 — Mortality in Puerto Rico after Hurricane Maria
After surveying individuals and comparing against demographic and geographic information, the estimated Puerto Rican death count due to Hurricane Maria increases 70-fold. [ Read More ]
18 — Association between physical exercise and mental health in 1·2 million individuals in the USA between 2011 and 2015: a cross-sectional study
Exercising regularly has an enormous positive impact on self-reported mental health. [ Read More ]
17 — Dynamic density shaping of photokinetic E. coli
Genetically engineered E. coli can form complex images in response to light stimuli. [ Read More ]
16 — Generating perfusion maps from structural optical coherence tomography with artificial intelligence
A machine learning approach can generate flow maps of a retina from anatomical OCT imagery. [ Read More ]
15 — The C. elegans AWA Olfactory Neuron Fires Calcium-Mediated All-or-None Action Potentials
The C. elegans connectome exhibits spiking action potential behavior despite the expectation that it exhibits analog neuron responses only. [ Read More ]
14 — Cortical Column and Whole Brain Imaging of Neural Circuits with Molecular Contrast and Nanoscale Resolution
Expansion microscopy, paired with lattice light sheet microscopy, can be used to image biological microstructure with lower spatial-acuity technologies such as light microscopy. [ Read More ]
13 — Clinically applicable deep learning for diagnosis and referral in retinal disease
Separation of segmentation and diagnosis of OCT images in a clinical setting improves human-machine cooperation. [ Read More ]
12 — Scene through the eyes of an apex predator: a comparative analysis of the shark visual system
Learning about shark eyes, retinas, and brains can teach us about early vertebrate evolution. [ Read More ]
11 — Hyperbrain network properties of guitarists playing in quartet
Musicians' brains appear to synchronize when they play music together. [ Read More ]
10 — A Complete Electron Microscopy Volume of the Brain of Adult Drosophila melanogaster
An entire fly brain is imaged at synaptic resolution, enabling large scale connectomics and neuroscience. [ Read More ]
09 — Thermal Imaging Reveals Changes in Body Surface Temperatures of Blacktip Sharks (Carcharhinus limbatus) during Air Exposure
Shark surface temperature increases steadily after a shark has been landed, which can affect the health of the shark as well as the shark's ecosystem. [ Read More ]
08 — An unusual association of hadrosaur and therizinosaur tracks within Late Cretaceous rocks of Denali National Park, Alaska
The first known colocated tracks of therizinosaurid and hadrosaurid species was found in Denali National Park. [ Read More ]
07 — A Distance-Dependent Distribution of Presynaptic Boutons Tunes Frequency-Dependent Dendritic Integration
Synapses shrink in size the further they are from the soma of a CA1 pyramidal cell, but the role of short-term facilitation increases. [ Read More ]
06 — New information on the cranial morphology of Avimimus (Theropoda: Oviraptorosauria)
New specimens of oviraptorosaurians demonstrate previously unconfirmed dissimilarities between the skull morphology of birds and Avivmimus dinosaurs. [ Read More ]
05 — The genetic basis of diurnal preference in Drosophila melanogaster
Though crepuscular in the lab, fruit flies can be both nocturnal or diurnal, and by selecting for one or the other artificially, we can identify the genes responsible for phase preference changes. [ Read More ]
04 — Single-cell molecular connectomics of intracortically-projecting neurons
Combining MAPseq with RNA sequencing, ConnectID enables researchers to overlay both transcriptome as well as and connectome information in cortical tissue. [ Read More ]
03 — All-Optical Machine Learning Using Diffractive Deep Neural Networks
Diffracting real light rather than routing digital neurons makes this 3D neural network run at the speed of light. [ Read More ]
02 — Single neurons may encode simultaneous stimuli by switching between activity patterns
In the presence of two stimuli, a neuron may fluctuate between response patterns to convey its responses to both stimuli. [ Read More ]
01 — Reevaluation of ‘endocostal ossifications’ on the Kebara 2 Neanderthal ribs
The Kebara 2 Neanderthal skeleton's abnormal rib ossifications are likely the result of a simple genetic disorder, and did not greatly impact day to day life. [ Read More ]
31 — On-the-Fly Power-Aware Rendering
Modulating the resources allocated to different stages of the render process optimizes for render quality without sacrificing power consumption budgets. [ Read More ]
30 — Detecting Influence Campaigns in Social Networks Using the Ising Model
Taking inspiration from physics simulations, it is possible to identify botnets on Twitter by their conspicuous behavior. [ Read More ]
29 — High-frequency observations from a deep-sea cabled observatory revealseasonal overwintering of Neocalanus spp. in Barkley Canyon, NE Pacific:Insights into particulate organic carbon flux
Copepod migration is more easily monitored using video feeds and Doppler acoustic data from a powerful undersea observatory. [ Read More ]
28 — Scutoids are a geometrical solution to three-dimensional packing of epithelia
A novel solid shape best approximates the geometry of epithelial columnar cells along surfaces of complex curvature. [ Read More ]
27 — Seasonal migration of marsupial megafauna in Pleistocene Sahul (Australia–New Guinea)
Diprotodon dentition show evidence of latitudinal migration over the course of the animal's lifespan. [ Read More ]
26 — Protecting Intellectual Property of Deep Neural Networks with Watermarking
Training neural networks on out-of-domain inputs allows an author to watermark their model to prevent or detect theft. [ Read More ]
25 — Radar evidence of subglacial liquid water on Mars
Dielectric permittivity measurements taken from the south pole of Mars strongly suggest a liquid water lake beneath the polar glaciers. [ Read More ]
23 — Electron ptychography of 2D materials to deep sub-ångström resolution
A much higher electron microscopy resolution can be achieved by leveraging ptychography of diffracted low-energy electrons. [ Read More ]
22 — BB-Graph: A New Subgraph Isomorphism Algorithm for Efficiently Querying Big Graph Databases
BB-Graph is a new subisomorphism search algorithm that improves upon GraphQL and Cypher implementations. [ Read More ]
21 — Adversarial Reprogramming of Neural Networks
It is possible to construct and adversarial program that hijacks the compute power of a neural network and repurposes it to solve a different task. [ Read More ]
20 — Sampling Neuron Morphologies
A generative model based upon Markov chain monte carlo simulations creates realistic neuron morphologies. [ Read More ]
19 — Learning cellular morphology with neural networks
Networks fed 2D renders of neuron segmentation fragments were able to capture information about the morphology and biology of the neurons. [ Read More ]
18 — Injury Incidence and Patterns Among Dutch CrossFit Athletes
The authors conducted a survey of CrossFit athletes in the Netherlands to better understand injury incidence rate. [ Read More ]
17 — Genome-culture coevolution promotes rapid divergence of killer whale ecotypes
Orcas' quick spread across the planet was facilitated both by social knowledge transfer as well as genetic drift. [ Read More ]
16 — Female hurricanes are deadlier than male hurricanes
Hurricanes with female names are less likely to elicit preparatory action than hurricanes of the same magnitude with male names. [ Read More ]
15 — Individual, unit and vocal clan level identity cues in sperm whale codas
Sperm whale codas covary with geography and social clan, reinforcing the social complexity hypothesis. [ Read More ]
14 — Anatomy of an Asteroid Break-Up: The Case of P/2013 R3
Under YORP-induced spin, an asteroid breaks apart over the course of several months in 2013-14. [ Read More ]
12 — New ootype prismatoolithids from the Late Cretaceous, Laiyang Basin and its significance
Newly found fossilized eggs teach us a great deal about dinosaur diversity and distribution. [ Read More ]
11 — The pitch of babies’ cries predicts their voice pitch at age 5
The pitch of a baby's cries and the ratio of its finger lengths is predictive of voice fundamental frequency later in life. [ Read More ]
10 — Testing the AC/DC hypothesis: Rock and roll is noise pollution and weakens a trophic cascade
Loud music or urban sounds can have dramatic impacts on the trophic webs of invertebrate/plant ecosystems. [ Read More ]
09 — Electric Fields Elicit Ballooning in Spiders
Spiders use the electric potential of the atmosphere in order to take off and fly to new locations. [ Read More ]
08 — Mangrove vegetation as a protective solution for coastal protections
Mangrove roots provide a strong barrier in preventing high speed turbulent flow in flood areas. [ Read More ]
07 — Syntax Error Recovery in Parsing Expression Grammars
This modified PEG uses a separate grammar in order to recover from errors. [ Read More ]
06 — Beef Cattle Instance Segmentation Using Fully Convolutional Neural Network
This simple neural network can segment individual animals in a cattle farm. [ Read More ]
05 — Human-level performance in first-person multiplayer games with population-based deep reinforcement learning
Agents trained to collaboratively play capture-the-flag video games outperformed human teams. [ Read More ]
04 — A Hidden Markov Model to Estimate the Time Dairy Cows Spend in Feeder Based on Indoor Positioning Data
A Markov model predicts cow feeding behavior based upon position information with extremely high accuracy. [ Read More ]
03 — Evaluating Fitness to Perform in Surgical Residents after Night Shifts and Alcohol Intoxication: The development of a “Fit-to-Perform” test
Surgical residents performed about as well on a battery of alertness tests after 14 hour shifts as they did at the legal BAC limit. [ Read More ]
02 — Lattice system of functionally distinct cell types in the neocortex
Rodent neocortex layer-5 contains long-range projection neurons which appear to be surrounded by a small column of atomic cortical computation. [ Read More ]
01 — A comparative study of artificial intelligence and human doctors for the purpose of triage and diagnosis
Babylon AI is a probablistic graph model that aims to diagnose clinical conditions as well as human physicians. [ Read More ]
30 — Learning Social Conventions in Markov Games
By including the observation of other agents in the self-play AI model, new agent strategies follow societal convention. [ Read More ]
29 — Modelling the brain response to arbitrary visual stimulation patterns for a flexible high-speed BCI
This work proposes a bidirectional technique for converting between EEG and stimulation pattern timeseries data to aid in BCI control. [ Read More ]
28 — Comparison of Hospital Mortality and Readmission Rates for Medicare Patients Treated by Male vs Female Physicians
Female physicians' patients tended to have better healthcare outcomes than male physicians'. [ Read More ]
27 — Important juvenile manta ray habitat at Flower Garden Banks National Marine Sanctuary in the northwestern Gulf of Mexico
Little is known about juvenile oceanic manta rays, but a new region of the Gulf of Mexico may be a recurring nursery-like habitat. [ Read More ]
26 — Active vision varies across the cardiac cycle
Self-paced subjects preferred to encounter new stimuli during systole than diastole. [ Read More ]
25 — World Models
This tripartite agent model learns to play games in the form of neural network hallucinations [ Read More ]
24 — Activity seascapes highlight central place foraging strategies in marine predators that never stop swimming
Unlike most CPF animals, sharks require a modified behavioral model because they never stop swimming. [ Read More ]
23 — Discovery of a new song mode in Drosophila reveals hidden structure in the sensory and neural drivers of behavior
Applying a computational approach categorizes fruit fly mating songs and the neural underpinnings of these different singing modes better than existing state of the art. [ Read More ]
22 — The Only Known Jawed Vertebrate with Four Eyes and the Bauplan of the Pineal Complex
The first ever known case of a vertebrate with four 'eyes' changes the evolutionary model of the pineal organ. [ Read More ]
21 — Prosthesis with neuromorphic multilayered e-dermis perceives touch and pain
This prosthetic 'skin' can communicate tactile sensation to the wearer of a prosthetic device. [ Read More ]
20 — A Machine Learning Approach to Discover Rules for Expressive Performance Actions in Jazz Guitar Music
Aligning fake-book lead sheets to commercial musical recordings enables this ML algorithm to learn common ornamentation techniques for jazz music. [ Read More ]
19 — Aqueous-based tissue clearing in crustaceans
This decalcification and bleaching process pretreats exoskeleton-bearing creatures for tissue-clearing processes like CUBIC. [ Read More ]
18 — Neural scene representation and rendering
GQNs learn to represent 3D scenes and can predict their layout based upon 2D images. [ Read More ]
16 — A Connectome Based Hexagonal Lattice Convolutional Network Model of the Drosophila Visual System
it is possible to reconstruct a working computer vision neural network from data derived from scans of the fruit fly visual system. [ Read More ]
15 — Late Ediacaran trackways produced by bilaterian animals with paired appendages
Fossilized tracks in China are likely the first ever evidence of paired appendages in a bilaterian in the fossil record. [ Read More ]
14 — CrisisMMD: Multimodal Twitter Datasets from Natural Disasters
The CrisisMMD dataset provides Twitter-based imagery and accompanying metadata to improve the humanitarian response to worldwide disasters. [ Read More ]
13 — Dank Learning: Generating Memes Using Deep Neural Networks
A neural network generates internet memes and their captions. [ Read More ]
12 — A Scalable Framework for Trajectory Prediction
A Markov model based system can predict the complex trajectory of a vehicle moving along a road network. [ Read More ]
11 — Why do deep convolutional networks generalize so poorly to small image transformations?
Common classification CNNs are sensitive to even very slight changes in a target image due to inherent biases in almost all commonly used training datasets. [ Read More ]
10 — A Process for Capturing CO2 from the Atmosphere
This direct-air-capture technique pulls carbon dioxide from the air at a cost of about 100-200 USD per ton. [ Read More ]
09 — The logic of single-cell projections from visual cortex
Using MAPseq, it is possible to locate and characterize the downstream projections from visual cortex in a whole mouse brain. [ Read More ]
08 — Predicting oral disintegrating tablet formulations by neural network techniques
This neural network can predict the disintigration time of dissolvable pharmaceutical tablets. [ Read More ]
07 — Caffeine dosing strategies to optimize alertness during sleep loss
By optimizing caffeine doses and timing, it is possible to reduce the total amount of caffeine required to maintain neurobehavioral performance on a battery of tests. [ Read More ]
06 — Playing Atari with Six Neurons
A tiny neural net is big enough to play certain Atari games when game-state is provided in a simplified form. [ Read More ]
05 — Rational Temporal Predictions Can Underlie Apparent Failures to Delay Gratification
This 2013 reproduction of the marshmallow test sought to understand how accurate and inaccurate predictions of time-delays influenced the seemingly irrational choices behind children's behavior. [ Read More ]
03 — Biodiversity on Indigenous lands equals that in protected areas
Indigenous lands have wildlife biodiversity levels that are on par with worldwide protected lands. [ Read More ]
02 — Cyclist’ safety perceptions of cycling infrastructure at un-signalised intersections: Cross-sectional survey of Queensland cyclists
Cyclists paradoxically prefer less convenient urban design, such as intersections at which they must cede right-of-way to motorists. [ Read More ]
01 — KG²: Learning to Reason Science Exam Questions with Contextual Knowledge Graph Embeddings
A knowledge graph embedding is learned by a neural network in order to answer complex exam-style questions based upon knowledge from a large corpus. [ Read More ]
31 — The Morphology and Circuity of Walkable and Drivable Street Networks
OSMnx, a python package, can be used to determine circuity, a measure of walkability versus drivability in US cities. [ Read More ]
30 — Deep Reinforcement Learning in Ice Hockey for Context-Aware Player Evaluation
This deep learning approach evaluates hockey players based upon their total impact on the outcome of a game. [ Read More ]
29 — Revisiting the Marshmallow Test: A Conceptual Replication Investigating Links Between Early Delay of Gratification and Later Outcomes
The famous marshmallow test was re-run using a larger population and better controls, and found less of a correlation with delayed gratification and later outcomes. [ Read More ]
28 — Deep Watershed Detector for Music Object Recognition
This deep-learning approach to digitizing hand-writing musical notation uses a variant of the common watershed segmentation approach. [ Read More ]
27 — Mode-Adaptive Neural Networks for Quadruped Motion Control
This deep learning approach automates the animation of complex quadruped navigation and locomotion over arbitrary paths and speeds in 3D. [ Read More ]
26 — A Burmese amber tick wrapped in spider silk
A tick fossilized in amber is wrapped with what appears to be spider silk, demonstrating a rarely-seen interaction between spiders and ticks in the fossil record. [ Read More ]
25 — Fast animal pose estimation using deep neural networks
This pose-estimation and animal tracking software uses deep learning to estimate the positions and configurations of a freely behaving animal. [ Read More ]
24 — The Roles of Supervised Machine Learning in Systems Neuroscience
Machine learning has become an irreplaceable tool in the neuroscientist's toolbox for a number of distinct but co-emergent reasons. [ Read More ]
23 — Estimating the impact of structural directionality: How reliable are undirected connectomes?
Taking directed brain graphs and removing directionality to simulate non-invasive data-recording techniques removes meaningful signal from the data, suggesting that directionality is a core component of understanding a brain graph. [ Read More ]
22 — Predicting Electricity Outages Caused by Convective Storms
To train machine-learning approaches to predict the power-grid damage caused by storms, it is necessary to artificially over-sample rarer events. [ Read More ]
21 — Organic matter loading by hippopotami causes subsidy overload resulting in downstream hypoxia and fish kills
Hippopotami poop a lot in the same place and when it all washes away it kills all the fish. [ Read More ]
20 — Evidence of a plume on Europa from Galileo magnetic and plasma wave signatures
New analyses of decade-old data from the Galileo mission around Jupiter suggest that the space probe passed through a plume of Europa. [ Read More ]
19 — Exploring the Color of Glass: Letters of Recommendation for Female and Male Medical Faculty
Recommendation letters for women pale in comparison with the level of focus and emphasis behind letters written for men. [ Read More ]
18 — Gene Duplication and Gene Expression Changes Play a Role in the Evolution of Candidate Pollen Feeding Genes in Heliconius Butterflies
Pollen-feeding butterflies upregulate cocoonase genes in their proboscis, entirely separate from where most digestive enzymes are produced in butterflies. [ Read More ]
17 — Comprehensive machine learning analysis of Hydra behavior reveals a stable basal behavioral repertoire
Via computer vision fed into unsupervised machine learning, video feeds of the cnidarian hydra vulgaris can be categorized into highly stereotyped behaviors that shed light on the fundamental capabilities of even small systems of neurons. [ Read More ]
16 — The structure factor of primes
Armed with techniques conventionally used in crystallography, it is possible to run simulations that characterize the fractal structure of prime number distribution along the number line. [ Read More ]
15 — Unsupervised Intuitive Physics from Visual Observations
This unsupervised learning approach builds intuitions about physics using only top-down video feeds, entirely without simulators. [ Read More ]
14 — Tweet success? Scientific communication correlates with increased citations in Ecology and Conservation
Altmetrics and Twitter success metrics correlate with higher citation counts in ecology and conservation studies. [ Read More ]
13 — Automatic discovery of cell types and microcircuitry from neural connectomics
Using neuroscientific priors, this probablistic technique can cluster neurons in a connectome by hypothesized cell type. [ Read More ]
12 — Classification of Household Materials via Spectroscopy
Using spectroscopy, a robot can predict the material composition of its surroundings with great accuracy. [ Read More ]
11 — Jupiter’s atmospheric jet streams extend thousands of kilometres deep
Measuring the gravitational field of Jupiter enables us to understand its dynamic atmosphere, the depth of jet-streams, and the mass ratio of the dynamical atmosphere to its gas core. [ Read More ]
10 — Solving Sudoku with Ant Colony Optimisation
Ant-colony optimization is useful to solve the NP-hard challenge of completing a sudoku. [ Read More ]
09 — Detecting Mutations by eBWT
Leveraging a form of the Burrows-Wheeler transform, it is possible to indentify SNP clusters in a human genome. [ Read More ]
08 — Eat & Tell: A Randomized Trial of Random-Loss Incentive to Increase Dietary Self-Tracking Compliance
Individuals who were penalized a random penalty for food-diary non-compliance were more likely to comply than individuals who where penalized a consistent, known amount. [ Read More ]
07 — Casual 3D Photography
A reconstruction algorithm converts a casually recorded set of photographs into a semi-3D scene mesh. [ Read More ]
06 — Global Catfish Biodiversity
One in every 20 species of vertebrates is a catfish, and understanding this diverse and ubiquitous fish sheds light on ecological and biological processes. [ Read More ]
05 — Whole-genome sequencing of the blue whale and other rorquals finds signatures for introgressive gene flow
Many types of rorqual whales share a networked evolutionary tree, and hippopotamuses are not whales. [ Read More ]
04 — Dynamic Structural Similarity on Graphs
Dynamic Structural Similarity is a node-to-node similarity measure that can be computed iteratively in order to refine a result. [ Read More ]
03 — The readability of scientific texts is decreasing over time
The readability of academic papers has been decreasing rapidly since the late 1880s. [ Read More ]
02 — Spontaneous behaviors drive multidimensional, brain-wide population activity
Background chatter characteristic of sensory cortices receiving not sensory inputs can likely be explained by the neural expression of the simultaneous occurrence of behaviors. [ Read More ]
01 — The Best of Both Worlds: Combining Recent Advances in Neural Machine Translation
Combining existing machine translation neural networks, Google proposes a hybrid system which outperforms its ancestors. [ Read More ]
30 — Getting to the bottom of anal evolution
The evolutionary origins of the anus are hugely impactful in understanding the organization of the animal world, but it's also hotly debated. [ Read More ]
29 — Synaptic circuits and their variations within different columns in the visual system of Drosophila
Reconstructing seven columns of fly medulla enables us to understand and differentiate between reconstruction errors and natural variations, and also provides a substrate for more principled learning about species stereotypy. [ Read More ]
28 — Observing the cell in its native state: Imaging subcellular dynamics in multicellular organisms
This novel microscopy technique combines light sheet microscopy with adaptive optics to achieve high spatiotemporal resolution in-vivo. [ Read More ]
27 — Of flying frogs and levitrons
Using a strong enough magnet, it is possible to non-destructively levitate water, and water-based creatures. [ Read More ]
26 — A carbon nanotube tape for serial-section electron microscopy of brain ultrastructure
Carbon nanotubes form a better substrate for serial-section electon microscopy than the conventional Kapton tape. [ Read More ]
25 — MuseGAN: Multi-track Sequential Generative Adversarial Networks for Symbolic Music Generation and Accompaniment
MuseGAN uses a variety of models to approximate human-quality musical composition using GAN architectures. [ Read More ]
24 — Application of smoothed particle hydrodynamics to modeling mechanisms of biological tissue
Sibernetic is a physics engine framework for simulating fluids of various viscosities and spring coefficients over time. [ Read More ]
23 — Fast and simple comparison of semi-structured data, with emphasis on electronic health records
This patient-fingerprinting method provides an anonymized, HIPAA-compliant way of determining patient similarity, without exposing sensitive ePHI. [ Read More ]
22 — Progressive Growing of GANs for Improved Quality, Stability, and Variation
NVIDIA researchers propose a progressively-refined image-generation technique to improve GAN image synthesis, and propose a more sensitive metric of GAN performance. [ Read More ]
21 — Gating and control of primary visual cortex by pulvinar
Lateral pulvinar gates LGN inputs to primary visual cortex in primates, permitting or disallowing inputs to reach the visual processing stream. [ Read More ]
20 — A generative vision model that trains with high data efficiency and breaks text-based CAPTCHAs
Recursive Cortical Networks are a novel ML architecture that are modeled closely after visual cortex in order to better parse complex scenes such as CAPTCHAs. [ Read More ]
19 — Dynamic representation of 3D auditory space in the midbrain of the free-flying echolocating bat
Free-flying brown bats encode 3D movement and location in the sensory cells of the superior colliculus. [ Read More ]
18 — The functional organization of cortical feedback inputs to primary visual cortex
LM feedback inputs to mouse V1 are largely retinotopic, but also include some distal feedback connections that violate a strict retinotopic map. [ Read More ]
17 — Automated diagnosis of colon cancer using hyperspectral sensing
Hyperspectral spectroscopy enables surgeons to identify cancerous tumors in colorectal cancer patients even when they are not visible to the naked eye. [ Read More ]
16 — Transductive Adversarial Networks (TAN)
Transductive adversarial networks enable training on labelled data in one domain and applying the model to unlabelled data in another domain, improving the feasibility of machine learning in restricted data ecosystems. [ Read More ]
15 — Voices Obscured in Complex Environmental Settings (VOICES) corpus
The VOICES dataset consists of vocal utterances from different positions in a room, along with distracting factors like radio, TV, or music. [ Read More ]
14 — CubeNet: Equivariance to 3D Rotation and Translation
CubeNet is a novel type of CNN that exhibits invariance to right-angle rotations of the target input in 3D space. [ Read More ]
13 — STAIR Actions: A Video Dataset of Everyday Home Actions
The STAIR Actions dataset records over 100,000 five-second videos in order to represent common activities that take place in the home. [ Read More ]
12 — Segmentation of Multiple Sclerosis lesion in brain MR images using Fuzzy C-Means
Fuzzy-c-means clustering alongside canny edge detection provides a fast MS lesion segmentation algorithm. [ Read More ]
11 — Discovering Hidden Factors of Variation in Deep Networks
Asking a classifier to output what it considers the "extremes" of its labels demonstrates that nets store a continuum even when asked just for a label. [ Read More ]
10 — Functional Organization and Dynamic Activity in the Superior Colliculus of the Echolocating Bat, Eptesicus fuscus
The brown bat superior colliculus contains specialized machinery that interleaves motor and sensory neurons for high-speed transfer of sensory and motor information between the systems. [ Read More ]
09 — Anxiety Cells in a Hippocampal-Hypothalamic Circuit
Fear and avoidance are controlled by different projections from ventral hippocampus, where anxiogenic stimuli are projected to lateral hypothalamus and basal amygdala. [ Read More ]
08 — Interbrain cortical synchronization encodes multiple aspects of social interactions in monkey pairs
When one monkey watches another monkey experience a change in movement and location, it too experiences a similar-looking configuration of signals in its primary motor cortex. [ Read More ]
07 — The growth pattern of Neandertals, reconstructed from a juvenile skeleton from El Sidrón (Spain)
Neanderthal skeletal growth patterns very closely mirror those of modern humans, though vertebral fusion occurred much later. [ Read More ]
06 — Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs
Recurrent Conditional GANs can be used to generate not only images, but also time-series data that, in this case, are used to simulate medical data. [ Read More ]
05 — Gender Bias in Sharenting: Both Men and Women Mention Sons More Often Than Daughters on Social Media
Parents tend to more often share and like posts about sons than about daughters on the VK Russian social media website. [ Read More ]
04 — AxonDeepSeg: automatic axon and myelin segmentation from microscopy data using convolutional neural networks
AxonDeepSeg is a deep U-net neural network that segments axons and myelin from lower-resolution electron microscopy data. [ Read More ]
03 — A gate-and-switch model for head orientation behaviors in C. elegans
The RIA neurons of the C. elegans nematode integrate sensory inputs with motor feedback in order to generate complex posture-dependent behaviors. [ Read More ]
02 — Electron Microscopic Reconstruction of Functionally Identified Cells in a Neural Integrator
Overlaying light microscopy with electron microscopy enables us to understand both the neuron classifications as well as the anatomy of a population of neurons. [ Read More ]
01 — MRI2MRI: A deep convolutional network that accurately transforms between brain MRI contrasts
MRI2MRI is a tool that transforms MRI imagery between MRI modalities using style-transfer based nets. [ Read More ]
31 — Monkeys use the rod‐dense retinal region rather than the fovea to visually fixate small targets in scotopic vision
In dark conditions, rhesus monkeys look slightly away from a stimulus in order to best place it inside their retinas' high rod density regions. [ Read More ]
30 — A longitudinal study of phenotypic changes in early domestication of house mice
After only 20 generations, domestic mice differ from their fully wild relatives in a variety of ways, including snout size, brain volume, and fur coloration. [ Read More ]
29 — Text2Shape: Generating Shapes from Natural Language by Learning Joint Embeddings
Using a joint learned embedding, Text2Shape can convert natural language descriptions into 3D representations of furniture or other objects. [ Read More ]
28 — Moving magnetoencephalography towards real-world applications with a wearable system
Leveraging the atomic behavior of Rubidium vapor, it is possible to build a mobile, less energy-hungry MEG scanner that is helmet-mounted instead of stationary. [ Read More ]
27 — The connection-set algebra — a novel formalism for the representation of connectivity structure in neuronal network models.
A formalized grammar enables us to speak about biological neuronal network approximations with mathematical rigor. [ Read More ]
26 — Tell Me Why Is It So? Explaining Knowledge Graph Relationships by Finding Descriptive Support Passages
This system designed by IBM and IIT searches a corpus of text for the most relevant phrases to explain a given entity triple from a knowledge graph. [ Read More ]
25 — The Circuit Motif as a Conceptual Tool for Multilevel Neuroscience
Characterizing individual neuronal circuit motifs is a powerful way of understanding the computational primitives of the brain. [ Read More ]
24 — A single spectrum of neuronal identities across thalamus
The thalamus acts as a filtering and processing step in the pipeline of sensory information entering cortex. [ Read More ]
23 — Limits of Predictability in Human Mobility
More than ninety percent of human mobility is potentially predictable based upon 50,000 users' mobile phone tracking data. [ Read More ]
22 — Neural Network Quine
This proposed method for designing a self-replicating neural network implements the concept of a code-less quine. [ Read More ]
21 — Breathing Life Into Dinosaurs: Tackling Challenges of Soft‐Tissue Restoration and Nasal Airflow in Extinct Species
Using 3D reconstructions of bone from Pachycephalosaurids, it is possible to partially reconstruct 3D soft-tissue nasal-passage geometry. [ Read More ]
20 — Equation-oriented specification of neural models for simulations
Brian2 uses membrane potential equation-based models in order to simulate biological spiking neural networks through Python. [ Read More ]
19 — Mathematical model of brain tumour with glia-neuron interactions and chemotherapy treatment
This system of equations provides a simple model for glioma tumor growth and suppression. [ Read More ]
18 — Whole-brain serial-section electron microscopy in larval zebrafish
Using the latest-and-greatest connectomics workflows and technology, a full-brain connectome and projectome dataset is generated from a larval zebrafish. [ Read More ]
17 — idtracker.ai: Tracking all individuals in large collectives of unmarked animals
idtracker.ai is a deep-learning animal-tracking system that can track individuals over time despite individuals crossing or touching. [ Read More ]
16 — PixelNN: Example-based Image Synthesis
A CNN feeds data into a pixelwise nearest-neighbor network in order to generate high resolution, highly controllable synthetic images from scratch. [ Read More ]
15 — Clusters of cyclones encircling Jupiter’s poles
New imagery from the Juno probe allow us to see the enormous cyclone systems storming on Jupiter's north and south poles. [ Read More ]
14 — Brain and eyes of Kerygmachela reveal protocerebral ancestry of the panarthropod head
Preserved neural tissue from an early ancestor of arthropods explains the evolutionary history of the centralized arthropod brain and eyes. [ Read More ]
13 — An exceptional Devonian fish from Australia sheds light on tetrapod origins
The Gogonasus specimen found in Australia in 2006 begins to close the gap between fishes and amphibians like Tiktaalik. [ Read More ]
12 — Does the Mid-Atlantic Ridge affect the distribution of abyssal benthic crustaceans across the Atlantic Ocean?
Of the surveyed abyssal species of mobile organisms in the deep sea around the Vera Fracture Zone, few species cross the Mid-Atlantic Ridge, making the MAR a robust dispersal barrier. [ Read More ]
11 — Automatic Machine Knitting of 3D Meshes
A 3D toolkit converts arbitrary manifold, watertight meshes into knittable patterns that can be loaded onto an automated knitting machine. [ Read More ]
10 — Convergent sequence evolution between echolocating bats and dolphins
The genetic changes to one gene that enables echolocation in bats also convergently evolved in dolphins and certain types of toothed whales. [ Read More ]
09 — Anamorphic development and extended parental care in a 520 million-year-old stem-group euarthropod from China
The earliest-known instance of extended parental care is found in a 520-million-year-old euarthropod fossil in China. [ Read More ]
08 — Closed-loop control of a modular neuromorphic biohybrid
Coupling a biological culture of neurons with a neuromorphic computing processor enables resarchers to modify, control, and measure the inputs and outputs of the neural network. [ Read More ]
07 — Correlations of Climate and Plant Ecology to Leaf Size and Shape: Potential Proxies for the Fossil Record
Identifying correlations between leaf physiognomy and climate enables us to better understand the climate of the prehistoric world by looking at leaves in the fossil record. [ Read More ]
06 — A surge of light at the birth of a supernova
An amateur astronomer, while testing a new telescope setup, coincidentally caught the birth of a new supernova, SN 2016gkg. [ Read More ]
05 — Satellite imagery analysis for operational damage assessment in Emergency situations
Phrasing disaster-zone detection as a satellite image segmentation problem enables machine learning approaches to hasten emergency response efforts. [ Read More ]
04 — Ancient DNA Reveals Late Pleistocene Existence of Ostriches in Indian Sub-Continent
Ancient DNA, found in manmade eggshell beads in west India, confirm that ostriches were still extant in India by ~25-40kya. [ Read More ]
03 — Brain-Computer Interface meets ROS: A robotic approach to mentally drive telepresence robots
This EEG platform sends signals to ROS to drive a humanoid Pepper robot for hands-free telepresence. [ Read More ]
02 — Trapping or tethering stones (TS): A multifunctional device in the Pastoral Neolithic of the Sahara
Tethering stones were used for a variety of purposes over several thousand years of human history. [ Read More ]
01 — Deep Learning for Isotropic Super-Resolution from Non-isotropic 3D Electron Microscopy
Two deep-learning approaches recover missing data along the low-resolution axis of anisotropic 3D electron microscopy datasets. [ Read More ]
28 — Rapid and Semi-Automated Extraction of Neuronal Cell Bodies and Nuclei from Electron Microscopy Image Stacks
CellSeeker is an open-source software product that finds cell bodies in an electron microscopy image stack for further analysis, without deep learning. [ Read More ]
27 — Humans quickly learn to blink strategically in response to environmental task demands
When expecting a short-duration stimulus, humans are likely to subconsciously supress blinking, and then compensate afterwards, according to a simple probablistic model. [ Read More ]
26 — StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks
StackGAN can convert freeform text into an image that matches that description, using two GANs stacked on top of each other in order to best refine details. [ Read More ]
25 — Image2Mesh: A Learning Framework for Single Image 3D Reconstruction
Using object recognition infrastructure paired with a dataset of 3D models, it is possible to capture a 3D scene reconstruction from only a single image. [ Read More ]
24 — U-Th dating of carbonate crusts reveals Neandertal origin of Iberian cave art
Dating the "canvas" material of cave paintings shows that the artwork predates the arrival of modern humans in Iberian caves and that these pieces should instead be attributed to Neanderthal artists. [ Read More ]
23 — Optimized Bacteria are Environmental Prediction Engines
It is evolutionarily favorable for a bacterial population to intentionally express a locally sub-optimal phenotype in order to best optimize for reproduction should the environmental conditions change. [ Read More ]
22 — A rock engraving made by Neanderthals in Gibraltar
Neanderthal cave carvings in Gibraltar suggest that Neanderthals were able to express themselves with abstract art, a skill once thought to be exclusive to modern humans. [ Read More ]
21 — Neural Turing Machines
A fully differentiable neural network that possesses devoted memory resources enables a network to outperform conventional machine learning techniques and incorporate historical information in computation. [ Read More ]
20 — The random walk of cars and their collision probabilities with planets
You're more likely to be killed by a ground-based Tesla than an outer-space Tesla, unless you visit another planet or live a really, really long time. [ Read More ]
19 — Tackling the Minimal Superpermutation Problem
The minimal superpermutation of a six-character string is still unknown, but we do know both that it disproves our prior understanding of superpermutations and that we can find it using brute force solutions to a related TSP. [ Read More ]
18 — End-to-end differentiable learning of protein structure
A new type of neural network predicts 3D protein structure faster, better, and with less initial knowledge than every other existing method. [ Read More ]
17 — Scalable and accurate deep learning for electronic health records
Using a deep-learning approach, it is possible to predict medical events, such as readmission or condition diagnosis, in a hospital population with extremely high accuracy and some degree of explainability. [ Read More ]
16 — Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records
DeepPatient predicts patient outcomes by embedding patients in a feature-space that enables it to use prior cases in the EHR to identify similar patient stories. [ Read More ]
15 — Mutual inhibition of lateral inhibition: a network motif for an elementary computation in the brain
A simple graph structure exists in a variety of species for a variety of purposes, suggesting that this motif is a fundamental building block of neural circuitry. [ Read More ]
14 — Diurnal transcriptome atlas of a primate across major neural and peripheral tissues
Genes are often cyclically through the circadian cycle, and this study examines this cycle in baboons, compared to the lab mouse upon which many gene expression studies are based. [ Read More ]
13 — Deep Learning based Retinal OCT Segmentation
This automated segmentation approach to OCT retina imagery uses a fully connected convolutional neura network to improve upon existing state of the art algorithms. [ Read More ]
12 — Crowdsourcing biomedical research: leveraging communities as innovation engines
When charged with analyzing a large biomedical dataset, it is cost-effective to open the opportunity up to citizen scientists and benefit from crowd-sourced solutions. [ Read More ]
11 — A Feathered Dinosaur Tail with Primitive Plumage Trapped in Mid-Cretaceous Amber
This newly found amber-preserved specimen teaches us about Cretaceous dinosaur plumage, as well as the origins of modern avian anatomy. [ Read More ]
10 — Gliotransmission: Beyond Black-and-White
In this part of the Dual Perspectives series, it is argued that gliotransmission does take place in natural circumstances in the healthy brain. [ Read More ]
09 — Multiple Lines of Evidence Indicate That Gliotransmission Does Not Occur under Physiological Conditions
In this part of the Dual Perspectives series, it is argued that glia do not take part in the calcium-mediated release of neurotransmitters in order to signal information to neurons. [ Read More ]
08 — The Neuronal Gene Arc Encodes a Repurposed Retrotransposon Gag Protein that Mediates Intercellular RNA Transfer
Neurons can create capsid-like structures that enable them to send mRNA directly to neighboring cells, in a type of virus-like communication. [ Read More ]
07 — Highly dynamic fission–fusion species can exhibit leadership when traveling
Bottlenose dolphin groups in the Lower Florida Keys are conventionally led by a small group of delegated leaders, who know the area well and can lead the group to areas of resource richness and low danger. [ Read More ]
06 — Axonal synapse sorting in medial entorhinal cortex
Axons in mouse medial entorhinal cortex synapse onto inhibitory targets "sooner" along the length of the axon than onto excitatory targets, suggesting a complex signal-processing circuit. [ Read More ]
05 — Tau accumulations in the brains of woodpeckers
Woodpeckers, long assumed to be in some way protected from traumatic brain injury during pecking, are likely suffering more brain injury than previously thought. [ Read More ]
04 — Australia’s continental-scale acoustic tracking database and its automated quality control process
IMOS is a database that tracks animals around the coast of Australia using acoustic tracking techniques. The data are then sanitized using domain-specific knowledge about the speed and expected location of different species. [ Read More ]
03 — Imitation of novel conspecific and human speech sounds in the killer whale (Orcinus orca)
A female orca was trained to emulate human speech sounds and learned the skillset very quickly, suggesting that cetaceans have a highly developed vocal imitation ability. [ Read More ]
02 — PiaF: A Tool for Augmented Piano Performance Using Gesture Variation Following
PiaF enables a pianist to cue accompaniments using Kinect-recognized gestures. [ Read More ]
01 — Deep Predictive Models in Interactive Music
Musical interaction or co-improvization agents are now feasible due to advances in machine learning systems. [ Read More ]
31 — Medial preoptic circuit induces hunting-like actions to target objects and prey
The MPA-vPAG circuit, when active, increases the rate of nonsocial behavior, and acts as an on-switch for engagement with inanimate objects or prey. [ Read More ]
30 — Semantic attributes are encoded in human electrocorticographic signals during visual object recognition
Using data from electrode arrays implanted in the skulls of human subjects, it is possible to closely guess what sorts of objects an individual is looking at or thinking of based only on the magnitude and frequency of activation of certain brain areas. [ Read More ]
29 — Patterns of long-term climate variability and predation rates by a marine apex predator, the white shark Carcharodon carcharia
Great white shark predation on seals has a complex relationship with climate, weather, and environmental conditions. This dataset of hundreds of shark attacks on seals helps to start decoding this environment/predation relationship. [ Read More ]
28 — Mitochondria are physiologically maintained at close to 50 °C
Mitochondria in vitro appear to run up to 10ºC warmer than the rest of the cell due to the energy released through cellular respiration. [ Read More ]
27 — A photophoretic-trap volumetric display
3D volumetric displays can be generated by trapping a microscopic particle in a beam of light and quickly adjusting its position and color. [ Read More ]
26 — An LSTM Network for Highway Trajectory Prediction
An LSTM-driven model improves upon the state-of-the-art for predicting the position of other road vehicles alongside an autonomous vehicle up to ten seconds in the future. [ Read More ]
25 — The earliest modern humans outside Africa
A fossilized upper jaw found in Israel suggests that modern humans existed outside of Africa more than 50,000 years than previously thought. [ Read More ]
24 — The eardrums move when the eyes move: A multisensory effect on the mechanics of hearing
When some primates and humans move their eyes to fixate on a new location in space, the eardrums also move, adjusting in order to better resolve the origin of sounds and sights. [ Read More ]
23 — Regionalism in Disney Animation: Pink Elephants and Dumbo
The bizarre, drug-like "Pink Elephants" sequence in Dumbo wasn't some artistic fluke; it was the sign of a supressed art style shining through the Disney studios where it had been previously shunned. [ Read More ]
22 — TexT - Text Extractor Tool for Handwritten Document Transcription and Annotation
TexT is an interactive handwriting recognition engine that enables an annotator to annotate certain words and receive assistance from an AI assistant to find other instances of that same word. [ Read More ]
21 — Spot the Difference by Object Detection
This CNN identifies and locates suble changes to photographs of two very slightly different images, such as two versions of the same book, with minimal required human annotation. [ Read More ]
20 — Blind De-anonymization Attacks using Social Networks
Given an anonymized graph dataset of a social network, this system can use auxiliary datasets — such as other social networks — to reverse engineer identity. [ Read More ]
19 — Do Large (Magnitude ≥8) Global Earthquakes Occur on Preferred Days of the Calendar Year or Lunar Cycle?
Earthquakes cannot be predicted using the lunar cycle or the calendar day, which should be obvious but for some reason people still believe that junk. [ Read More ]
18 — Reinforcement Learning based Recommender System using Biclustering Technique
This recommender engine uses reinforcement learning to generate high-quality item recommendations for a purchasing user "buyer" agent. [ Read More ]
17 — Social Network based Short-Term Stock Trading System
This trading bot uses sentiment analysis pulled from tweets in order to determine which stocks to buy or sell, and when. [ Read More ]
16 — Building a Conversational Agent Overnight with Dialogue Self-Play
M2M provides a simple protocol for generating very large corpuses of human-like text from machine-generated outlines, toward the construction of natural language chatbots. [ Read More ]
15 — Do Dogs Know Related Rates Rather than Optimization?
Either dogs are good at mentally solving calculus problems, or their brains recompute local optima at every moment of a game of fetch, calculating an optimal path by approximation. [ Read More ]
14 — “Like Sheep Among Wolves”: Characterizing Hateful Users on Twitter
Hateful users can be identified by several interesting characteristics, including account-creation, followership ratios, and retweet networks. [ Read More ]
13 — Discovering the hidden community structure of public transportation networks
Community networks enable cities to perform analyses on the public transportation network, which in turn enables them to optimize the system, or model the spread of infectious disease across a city. [ Read More ]
12 — Moments in Time Dataset: one million videos for event understanding
The Moments in Time dataset records around one million video and audio clips of three-second actions, alongside human-annotated labels of what each video contains. [ Read More ]
11 — Discovery of a big void in Khufu’s Pyramid by observation of cosmic-ray muons
Using muon transmission imaging, it is possible to determine where there are spaces inside the Great Pyramid in Giza. [ Read More ]
10 — SemMedDB: a PubMed-scale repository of biomedical semantic predications
SemMedDB is a knowledge-graph representation of medical literature that has been automatically extracted from PubMed article listings and the cross-citations between them. [ Read More ]
09 — Multi-Objective Vehicle Routing Problem Applied to Large Scale Post Office Deliveries
An offline, multi-vehicle routing problem can be solved by overlaying a graph with a probability map of each destination's difficulty of access. [ Read More ]
08 — Spatial Information in a Non-retinotopic Visual Cortex
Though turtle primary visual cortex is not laid out like mammalian visual cortex, it appears to respond to high-level scene- or shape-based stimuli, suggesting a more holistic approach to vision in early vision areas in turtles. [ Read More ]
07 — Integration of grid maps in merged environments
When rats were introduced to two spaces, and then those spaces were connected, their grid-cell representation of the new connected space was a combination of old information about the discrete subspaces alongside new "reconciliation" grids that combined the two. [ Read More ]
06 — Identification of Pre-Existing Adaptive Immunity to Cas9 Proteins in Humans
Streptococcus and Staphylococcus are commonly used to generate Cas9 homologs for CRISPR gene therapies, but human patients may have immune reactions to these homologs that could result in ineffective or harmful responses. [ Read More ]
05 — How Will the Internet of Things Enable Augmented Personalized Health?
The future of healthcare resides in highly personalized systems such as wearable monitors, which will be integrated with more conventional electronic medical records to augment the existing patient records. [ Read More ]
04 — An Unsupervised Homogenization Pipeline for Clustering Similar Patients using Electronic Health Record Data
Clustering patient data in order to arrive at diagnoses is highly dependent on the feature reduction and normalization methods used to sanitize the dataset. [ Read More ]
03 — Lateral geniculate neurons projecting to primary visual cortex show ocular dominance plasticity in adult mice
Contrary to popular opinion which placed plasticity only in cortex, thalamus dLGN undergoes neuroplastic changes in response to ocular dominance in mice. [ Read More ]
02 — Disequilibrium in Gender Ratios among Authors who Contributed Equally
When authors are listed as having "contributed equally" to a paper, men are disproportionately listed first, regardless of alphabetical or random ordering that would guarantee actual equal listing. [ Read More ]
01 — Content-Aware Image Restoration: Pushing the Limits of Fluorescence Microscopy
The CARE network is a neural network that is able to supersample light-microscopy fluorescence images in order to dramatically refine the spatial and temporal resolution of the imagery while using a tiny fraction of the photon exposure required for conventional methods. [ Read More ]
31 — The Merits of Sharing a Ride
The NP-hard problem of ride-sharing is increased in complexity when multiple rider routes are introduced. This paper used a real-life dataset of taxi routes in NYC as a testing metric. [ Read More ]
30 — Microglia and macrophages in brain homeostasis and disease
Microglia and other CNS macrophages are responsible for much more than just simple structural or immune mediation in the brain, and further research into glial cells is likely to shed light on poorly-understood disorders like ALS, Alzheimer's, Parkinson's, and many more. [ Read More ]
29 — 50 Hz volumetric functional imaging with continuously adjustable depth of focus
Using a variable Bessel beam focus, it is possible to image living brain tissue in vivo at much higher spatial and temporal resolution than with existing two-photon techniques. [ Read More ]
28 — Chewing sandpaper: grit, plant apparency, and plant defense in sand-entrapping plants
Certain desert plants exude a sticky substance to maintain a sandy coating, which acts as a defense against small herbivores such as caterpillars which find this sand to be unpleasant to consume. [ Read More ]
27 — Eye lens radiocarbon reveals centuries of longevity in the Greenland shark (Somniosus microcephalus)
Using radiocarbon dating and benchmarks such as the onset of thermonuclear weapon testing, it can be demonstrated that the Greenland shark is one of the longest-living animals in the world. [ Read More ]
26 — A Sensorimotor Circuit in Mouse Cortex for Visual Flow Predictions
Mouse motor cortex and anterior cinculate cortex participate in circuits that predict visual flow of stimuli across visual cortex in response to endogenous movements. [ Read More ]
25 — 3D Axonal Morphometry of White Matter
By segmenting white matter axonal tracts, it is possible to see morphological differences associated with traumatic brain injury in rat corpus callosa. [ Read More ]
24 — Live birth in the Devonian period
This Materpiscis attenboroughi holotype is the first known evidence of vertebrate live birth, moving the earliest-known live birth back by 200 million years. [ Read More ]
23 — A Deep Learning Interpretable Classifier for Diabetic Retinopathy Disease Grading
This neural network emphasizes human explainability when reporting diagnostic level of diabetic retinopathies in the EyePACS retinal dataset. [ Read More ]
22 — Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions
Tacotron 2 reproduces human speech at a creepy-good level by converting text first to a spectrogram and then to a waveform, using two predictive generative neural networks. [ Read More ]
21 — Evolution of highly diverse forms of behavior in molluscs
Despite their phylogenetic similarity, different subtypes of mollusc evolved very different types of nervous systems and behavior. [ Read More ]
20 — HOLA: Human-like Orthogonal Network Layout
HOLA automatically lays out graphs in a 2D orthogonal view that maximizes human readability, improving interpretation and speed of parsing. [ Read More ]
19 — Whole genome sequencing in psychiatric disorders: the WGSPD consortium
The Whole Genome Sequencing for Psychiatric Disorders Consortium aims to provide data and analysis toolkits for more than 200 thousand individual full genomes in pursuit of genetic underpinnings of neuropsychiatric disorders. [ Read More ]
18 — Improving automated multiple sclerosis lesion segmentation with a cascaded 3D convolutional neural network approach
This algorithm is able to segment multiple sclerosis white matter lesions with accuracy unprecedented on the MICCAI2008 MS MRI dataset. [ Read More ]
17 — Embodied Organization of Octopus vulgaris Morphology, Vision, and Locomotion
Octopus motor systems are probablistic, using distributed and bottom-up contributions that contribute to a central goal-seeking model — rather than top-down, central motor programs common in vertebrates. [ Read More ]
16 — Theta and alpha oscillations are traveling waves in the human neocortex
Brain oscillations ripple across cortex like waves across the surface of a pond, signalling — or perhaps even recruiting — the high-level coordination of brain regions for goal-based actions. [ Read More ]
15 — Female grant applicants are equally successful when peer reviewers assess the science, but not when they assess the scientist
Surprising no one, women were awarded fewer grants by CIHR peer review processes than men. When controlling for age and experience, this was shown to be a function of gender-bias. [ Read More ]
14 — The cognitive map in humans: spatial navigation and beyond
The cognitive map, well-supported with electrophysiological data from decades of research, is also supported by a variety of other methods, including functional MRI scans in humans. [ Read More ]
13 — Force-Directed Edge Bundling for Graph Visualization
This method of edge-bundling improves the readability and interpretability of large, congested graphs, without sacrificing runtime speed or node visibility. [ Read More ]
12 — FlagIt: A System for Minimally Supervised Human Trafficking Indicator Mining
FlagIt is a scalable, streaming, automated text-processing pipeline that helps identify and track sex trafficking ads or websites on the internet and dark web. [ Read More ]
11 — Nintendo Super Smash Bros. Melee: An 'Untouchable' Agent
Using reinforcement-learning, the authors were able to develop an "untouchable" player of Super Smash Bros. Melee that could survive for several minutes even on the hardest AI settings. [ Read More ]
10 — Terrestrial effects of moderately nearby supernovae
Plants, animals, and the atmosphere of Earth itself might have all been affected millions of years ago by nearby stars going supernova and showering the earth with visible light and radiation. [ Read More ]
09 — Eye-Drops for Activation of DREADDs
Eye-drops are as good a method as the more conventional intraperitoneal injection for introducing clozapine-N-oxide to CNS neurons _or_ retinal ganglion cells. [ Read More ]
08 — An opening for humor in melancholy
Inflammatory molecules are more easily able to cross the blood brain barrier when a mouse is under stress conditions, causing depression-like symptoms as a result of circulating compounds. [ Read More ]
07 — Copulation in antiarch placoderms and the origin of gnathostome internal fertilization
Modern fish and amphibians, which reproduce using external fertilization, likely evolved from ancestors that reproduced utilizing internal fertilization, contrary to previously held hypotheses. [ Read More ]
06 — A Novel Brain Decoding Method: a Correlation Network Framework for Revealing Brain Connections
Using both topological and functional correlations between brain regions, it is possible to improve image-processing neural networks using bio-inspired features. [ Read More ]
05 — An interactive framework for whole-brain maps at cellular resolution
An open-source brain mapping visualization and annotation suite enables whole-brain neuroscientific study and online multi-team collaboration. [ Read More ]
04 — Simulated Annealing Algorithm for Graph Coloring
Using the Metropolis Algorithm Monte Carlo, a simulated annealing approach, it is possible to generate very good graph-colorings for arbitrarily large graphs. [ Read More ]
03 — Now Playing: Continuous low-power music recognition
This method from a Google research uses local processing and hardware to fingerprint music and identify it from a database of songs. [ Read More ]
02 — Knowledge Graph Embedding with Iterative Guidance from Soft Rules
Using the newly presented system, RUGE, it is possible to convert knowledge graphs to vector embeddings with much higher levels of accuracy than before. [ Read More ]
01 — High-Precision Automated Reconstruction of Neurons with Flood-filling Networks
Flood-filling networks are a new class of CNNs that use seed points to segment electron microscopy neuroscience data with an extraordinary level of accuracy. [ Read More ]
29 — Learning a neural response metric for retinal prosthesis
Using neural networks, it is possible to "learn" the optimal signals to send with a retinal prosthesis to most closely emulate the signals sent by the healthy retinal cells. [ Read More ]
28 — Spider silk reinforced by graphene or carbon nanotubes
If you feed spiders carbon nanotubes, their silk becomes even stronger and tougher, beating almost all manmade fibers. [ Read More ]
27 — The Structural Architecture of an Infectious Mammalian Prion Using Electron Cryomicroscopy
This work begins to dissect the 3D structure of prion proteins responsible for certain types of mammalian proteinopathies. [ Read More ]
26 — Speech Recognition for Medical Conversations
CTC and LAS models both generate practically useful medical transcriptions of audio conversation recordings between physicians and patients. [ Read More ]
25 — Project Lyra: Sending a Spacecraft to 1I/’Oumuamua (former A/2017 U1), the Interstellar Asteroid
'Oumuamua is the first interstellar object to ever be seen in the Solar System. Reaching it with a space probe for further study would stretch the limits of what current technology could achieve. [ Read More ]
24 — Allocation Problems in Ride-Sharing Platforms: Online Matching with Offline Reusable Resources
This novel model of OM-RR-KAD resource-matching accounts for dynamic pools of suppliers and demanders for allocation problems in ride-sharing or organ-donor type problems. [ Read More ]
23 — JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs
JamBot produces chord progressions that emulate common patterns found in human-generated music, but fails to produce repeating melodies or motifs. [ Read More ]
22 — Simulating optical coherence tomography for observing nerve activity: a finite difference time domain bi-dimensional model
The optical coherence tomography protocol, conventionally used as an anatomical measure, may be modifiable to record functional data as well. [ Read More ]
21 — Gene name errors are widespread in the scientific literature
Twelve years after the announcement that gene names were commonly mistaken for dates in supplemental materials generated by spreadsheet software like Excel, the errors are still prevalent in modern gene study literature. [ Read More ]
20 — Tsunami-driven rafting: Transoceanic species dispersal and implications for marine biogeography
The 2011 Japanse tsunami washed many foreign species to North American ecosystems, the implications of which may never be fully understood. [ Read More ]
19 — Whale swarm algorithm for function optimization
By emulating whales' hunting and communication patterns, this algorithm searches a large parameter-space quickly, arriving at local maxima quickly. [ Read More ]
18 — C. elegans discriminate colors without eyes or opsins
By altering the blue:amber ratio of the light reaching a foraging worm, it is possible to modulate the pyocyanin avoidance response. [ Read More ]
17 — CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning
This convolutional neural network boasts better-than-human performance in pneumonia identification from chest X-rays. [ Read More ]
16 — On the Complexity of Learning Neural Networks
Neural networks' computational complexity can be approximated using mathematical approaches. [ Read More ]
16 — Gerrymandering and Computational Redistricting
By generating an algorithm to impartially design voting districts in the United States, it's possible to generate better, more fair district lines that will improve voter impact. [ Read More ]
15 — Web Robot Detection in Academic Publishing
It is possible to detect which web viewers are real humans and which are automated bot site crawlers, which improves interest-tracking and viewership statistics in academic publication. [ Read More ]
14 — Open-World Knowledge Graph Completion
ConMask is a knowledge-graph completion algorithm that uses neural networks to parse natural langauge to understand complex relationships between named entities. [ Read More ]
13 — Worm-level Control through Search-based Reinforcement Learning
The tap-withdrawal reflex in C. elegans is similar to the control system needed to balance an inverted pendulum shape. [ Read More ]
12 — Neurodata Without Borders: Creating a Common Data Format for Neurophysiology
NWB is a recently proposed, specialized file format for consistently and interoperably storing neurophysiological information and metadata. [ Read More ]
11 — Data Breaches, Phishing, or Malware? Understanding the Risks of Stolen Credentials
Keyloggers and other malicious password exfiltration software are largely unchecked on the open black market of the internet, where password leaks coupled with some basic analysis can affect millions of people. [ Read More ]
10 — Rhythmical Photic Stimulation at Alpha Frequencies Produces Antidepressant-Like Effects in a Mouse Model of Depression
Photic light stimulation at alpha wave frequencies appeared to have antidepressant effects in CORT-induced depression models in mice. [ Read More ]
09 — The Ubiquity of Large Graphs and Surprising Challenges of Graph Processing: A User Survey
This study surveyed people who commonly used graphs in their everyday work and discovered some interesting facts about the use of graphs in research and commercial settings. [ Read More ]
08 — Six Challenges for Neural Machine Translation
These are six very tangible and quantifiable ways in which neural network machine translation systems suffer from the common afflictions of deep learning. [ Read More ]
07 — Simple Cortex: A Model of Cells in the Sensory Nervous System
This novel machine learning framework uses spiking neural networks and Numenta's hierarchical temporal memory to learn complex scenarios. [ Read More ]
06 — Framework for evaluation of sound event detection in web videos
This system provides a way to generate tags for a video using only sound features, which allows one to query a video collection for sound content as well as human-added tags. [ Read More ]
05 — The Axonal Cytoskeleton and the Assembly of Nodes of Ranvier
This review covers recent discoveries in the importance of axonal cytoskeleton architecture in the healthy development of a Node of Ranvier. [ Read More ]
04 — Embedding of Cortical Representations by the Superficial Patch System
The superficial patch system tiles across sensory cortex in a hexagonal pattern in many disparate species of mammal. This architecture may help us better understand the ways in which signals are shared across cortex both in sensory and non-sensory cortex. [ Read More ]
03 — Learning deep features for source color laser printer identification based on cascaded learning
This CNN architecture can identify by which of a population of laser printers a document was printed with very high accuracy. This has applications in forensics and document provenance. [ Read More ]
02 — Pigeon Perception of Letters of the Alphabet
Researchers studied the ways in which pigeons viewed English alphabet characters and discovered interesting patterns in which letters were most commonly confused. [ Read More ]
01 — Complementary action of chemical and electrical synapses to perception
Dynamic range is one metric that helps to explain the specific structure of cortical circuitry and why firing rates and branching distributions are the way that they are in biology. [ Read More ]
31 — Deep Spatial Regression Model for Image Crowd Counting
Using a CNN+LSTM network, it is possible to estimate the number of people in a photograph of a large crowd with a high degree of reliability. [ Read More ]
30 — Spiking Optical Flow for Event-based Sensors Using IBM’s TrueNorth Neurosynaptic System
Using TrueNorth spiking-neuron hardware and a retina-like camera, it is possible to design a direction-selective circuit that emulates the motion-sensitive neurons of a mammal. [ Read More ]
29 — One pixel attack for fooling deep neural networks
This paper presents a technique for fooling computer-vision algorithms into misclassifying images by only modifying a single pixel in color and intensity. [ Read More ]
28 — Point Neurons with Conductance-Based Synapses in the Neural Engineering Framework
The Neural Engineering Framework currently uses conductance-based neurons; these simulations can be improved by converting the synapses to be conductance-based as well. [ Read More ]
27 — FashionBrain Project: A Vision for Understanding Europe’s Fashion Data Universe
FashionBrain hopes to design intelligent fashion-prediction and modeling tools to improve the efficacy and time-to-market of the currently lower-tech fashion industry. [ Read More ]
26 — Lip2AudSpec: Speech reconstruction from silent lip movements video
This system provides a way to transcribe spoken text using only video of a speaking face with no accompanying audio. [ Read More ]
25 — Retinal Ganglion Cells and Circadian Rhythms in Alzheimer's Disease, Parkinson's Disease, and Beyond
Alzheimer's and Parinson's patients suffer from altered circadian rhythms. It is very possible that this is related to the commonly-seen degeneration of retinal ganglion cells. [ Read More ]
24 — FPGA-based ORB Feature Extraction for Real-Time Visual SLAM
Using custom hardware, it is possible to dramatically increase the framerate of a classic SLAM feature extraction algorithm. [ Read More ]
23 — Adversarial Networks for the Detection of Aggressive Prostate Cancer
A combination of GAN and semantic segmentation enables this algorithm to find prostate cancer in MRI scans of patients with aggressive tumors. [ Read More ]
22 — Data-driven approach for creating synthetic electronic medical records
Real patient records can be perturbed slightly to simulate a disease outbreak in a simulated clinical environment. [ Read More ]
21 — Connecting Harbours. A comparison of traffic networks across ancient and medieval Europe
This project registers first-millennium European harbors and ports as nodes in a directed multigraph to enable graph-analysis and interprtation of known transit or transport avenues. [ Read More ]
20 — Polar wildfires and conifer serotiny during the Cretaceous global hothouse
Using computed-tomography scans, it is possible to show that plants with fire-triggered seed-release already existed in the mid-Cretaceous. [ Read More ]
19 — Neurobehavioural Correlates of Obesity are Largely Heritable
Brain region thickness and size correlates with predisposition for obesity, and these traits are heritable, meaning that obesity may be tied intimately to neurobehavior and neurogenomics. [ Read More ]
18 — Identifying Designs from Incomplete, Fragmented Cultural Heritage Objects by Curve-Pattern Matching
A curve-matching algorithm helps reconstruct full Native American designs from shattered pottery using a combination of computer-vision and curve-analysis. [ Read More ]
17 — Resting-state connectivity predicts patient-specific effects of deep brain stimulation for Parkinson's disease
This is one proposed technique for finding the best brain-region targets for Parkinson's patient deep-brain-stimulation therapy. It uses fMRI data to target patient-specific brain areas, rather than prior, less tailored methods. [ Read More ]
16 — Skin Lesion Analysis Toward Melanoma Detection: A Challenge At The 2017 International Symposium On Biomedical Imaging (ISBI), Hosted By the International Skin Imaging Collaboration (ISIC)
Machine learning approaches dominated the 2017 ISBI competition for the accurate identification of melanoma skin tissue by image-data, but were far more data-hungry than their non-ML counterparts. [ Read More ]
15 — The inefficiency of Bitcoin revisited: a dynamic approach
By using a different time-series analysis technique to evaluate volatility and market-mean deviation, a strange destabilization event in 2014 becomes obvious in the Bitcoin trade market history. [ Read More ]
14 — CHIPS – A Service for Collecting, Organizing, Processing, and Sharing Medical Image Data in the Cloud
CHIPS is a platform for medical data management, sharing, and storage that uses containers to distribute workloads intelligently and securely. [ Read More ]
13 — Clusters of Driving Behavior from Observational Smartphone Data
This study followed 500 drivers to collect terabytes of data on the acceleration and speed of their driving style. This enabled the researchers to construct well-defined clusters of driving behavior. [ Read More ]
12 — Portable Tor Router: Easily Enabling Web Privacy for Consumers
A pocket-sized Tor Router enables users to connect multiple devices to a protected, Tor-enabled hotspot. [ Read More ]
11 — A Neural Clickbait Detection Engine: The Tuna Clickbait Detector at the Clickbait Challenge 2017
The technique presented here combines image evaluation with headline evaluation to determine if a post contains clickbait or not. [ Read More ]
10 — Machine Learning for Drug Overdose Surveillance
Using machine learning methods and public datasets, this paper characterizes and explains clusters of opioid overdoses in America. [ Read More ]
08 — Encoding of Tactile Context in the Mouse Visual Cortex
Mice combine whisker-sensation with visual signals in V1 — very early in the visual processing stream, changing the way the entire brain perceives the visual world. [ Read More ]
07 — Inference of topology and the nature of synapses, and the flow of information in neuronal networks
Determining the directionality of synapses in a graph-representation of a neural network is crucial to understand the circuitry of the brain. This paper shows a way to determine connectivity without direct access to anatomical data, using functional data instead. [ Read More ]
06 — Discovery of the Twitter Bursty Botnet
This paper demonstrates a method for detecting botnets on Twitter, and explains how this can give better ground-truth for automated detection systems. [ Read More ]
05 — ☎️👨⛵️🐋👌; or “Call me Ishmael” – How do you translate emoji?
This is an exploration and analysis of Emoji Dick, a project that translated Moby Dick into emoji. [ Read More ]
04 — Creating a Social Brain for Cooperative Connected Autonomous Vehicles: Issues and Challenges
A common language for autonomous vehicles can improve collaboration and information-sharing across a network fabric. [ Read More ]
03 — Adversarial Generation of Natural Language
Generating natural language is very hard. When using a GAN, it is still hard. [ Read More ]
02 — “Why Should I Trust You?” Explaining the Predictions of Any Classifier
LIME helps show humans what made the machine-learning algorithm pick the output it chose by demonstrating what input values were most important in making that decision. [ Read More ]
01 — Senenmut: an ancient Egyptian astronomer
Much can be learned about the astronomical findings of Senenmut, a friend and royal timekeeper-advisor to Pharaoh Hatshepsut, from the inscriptions on the ceiling of his tomb. [ Read More ]
30 — Competition and Success in the Meme Pool: a Case Study on Quickmeme.com
This paper characterizes the rise and fall of individual memes to determine if they can be modeled by the original Dawkins "meme" definition — and if so, how we can understand meme success. [ Read More ]
29 — OSMnx: New methods for acquiring, constructing, analyzing, and visualizing complex street networks
Here's a Python library for the generation of graphs from street and intersection data from OpenStreetMap. I love this library. [ Read More ]
28 — Optical Coherence Tomography (OCT) in Optic Neuritis and Multiple Sclerosis
OCT scanning technology is a powerful way to measure the progression of optic neuritis and multiple sclerosis. [ Read More ]
27 — A Discrete and Bounded Envy-free Cake Cutting Protocol for Any Number of Agents
How do you cut a cake in nths by eye without making someone jealous of someone else's slice? [ Read More ]
26 — Electrical synapses between inhibitory neurons shape the responses of principal neurons to transient inputs in the thalamus
Simulations confirm that electrical synapses between cells in mouse whisker thalamus have significant impact on the downstream spike-stream signals sent to cortex. [ Read More ]
25 — AAV-mediated CRISPR/Cas gene editing of retinal cells in vivo
This paper demonstrates one possible technique with which retinal genetic therapy can be performed on mammalian (living) eyes. [ Read More ]
24 — Yet Another ADNI Machine Learning Paper? Paving The Way Towards Fully-reproducible Research on Classification of Alzheimer’s Disease
Alzheimer's Disease datasets are growing in use and popularity, but the inhomogeneity of the data make reproducible science difficult. This paper is a step toward reproducible diagnosis. [ Read More ]
22 — PassGAN: A Deep Learning Approach for Password Guessing
Password guessers have always relied on dictionary size and simple heuristics. PassGAN makes more nuanced and accurate guesses based on password list leaks. [ Read More ]
21 — A Deep Structured Learning Approach Towards Automating Connectome Reconstruction from 3D Electron Micrographs
Segmentation might not be such a big unsolved problem in EM neuroscience connectomics after all. [ Read More ]
20 — Zinc for the treatment of the common cold: a systematic review and meta-analysis of randomized controlled trials
My mom always tells me to take zinc supplements when I have a cold and I always say there's no evidence to back that up, so here you go. [ Read More ]
19 — Musical NeuroPicks: a consumer-grade BCI for on-demand music streaming services
NeuroPicks is a way to use EEG signals from a human brain to customize streaming music to a user's tastes. [ Read More ]
18 — The microstructure of high frequency markets
The authors propose a mathematical model to track the wealth differential of a high-frequency trader acting on a market with non-zero friction. [ Read More ]
17 — Banjo Drum Physics — theoretical preliminaries
A banjo is a surprisingly complex family of oscillating systems that are difficult to model. This paper demonstrates a way to reduce the mathematical complexity of that model. [ Read More ]
16 — Aberrant cortical activity in multiple GCaMP6-expressing transgenic mouse lines
A genetic family of mice used commonly in neuroscience studies appears to suffer from epilepsy-like brain activity, which may affect the results of neuroscientific study. [ Read More ]
15 — Efficient codes and balanced networks
The brain is a closely intertwined network of inhibitory and excitatory signals. Understanding the relationship between the two will improve the way we understand systems neuroscience. [ Read More ]
14 — Microglia emerge as central players in brain disease
Microglia appear to play a significant role in the healthy and diseases states of central nervous tissue. Understanding this role can help us better understand brain disease and development. [ Read More ]
13 — Crows Rival Monkeys in Cognitive Capacity
Crows perform about as well on a working-memory task as rhesus monkeys and humans on "easy" levels, but underperform on more complex tasks. [ Read More ]
12 — Seeing faces is necessary for face-domain formation
The brain area responsible for monkeys' facial recognition fails to develop if the monkey cannot see faces; but instead, hand-recognition regions are strengthened. [ Read More ]
11 — Hybrid computing using a neural network with dynamic external memory
Using a neural network with accessible "RAM" enables more powerful and robust machine-learning algorithms that can incorporate more data and historical information in conclusions. [ Read More ]
10 — A Closer Look at Tetris: Analysis of a Variant Game
Using mathematical abstractions of Tetris gameplay, it is possible to prove that the game is unwinnable, and will eventually terminate in a loss. [ Read More ]
09 — Using Human Brain Activity to Guide Machine Learning
Using fMRI brain data to determine how easily humans categorize an image, neural networks' results are improved by punishing them more significantly when they fail on "easy" images. [ Read More ]
08 — Random walks for image segmentation
Image segmentation can be modeled as a graph-theory problem, using diffusion equations and electric-circuit equations to simulate a random walk across objects. [ Read More ]
07 — Adversarial Examples for Evaluating Reading Comprehension Systems
Slight perturbations to an input can dramatically damage the accuracy of a neural network. Here are some ways to test this susceptibility. [ Read More ]
06 — Comparative Analysis of Gene Expression for Convergent Evolution of Camera Eye Between Octopus and Human
A genome-wide study can start to explain why octopus eyes are really similar to human eyes, even though the species split long before the evolution of the camera-eye. [ Read More ]
05 — Large-Scale Simulations of Plastic Neural Networks on Neuromorphic Hardware
SpiNNaker is a new, neuromorphic computer that simulates millions of plastic synapses orders of magnitude more efficiently than the equivalent computations on conventional supercomputers. [ Read More ]
04 — Optical coherence tomography reflects brain atrophy in multiple sclerosis: A four-year study
OCT, a low-cost and simple scan, can often indicate the severity and extend of brain lesions in patients with multiple sclerosis, which can reduce the need for expensive brain-scans. [ Read More ]
03 — Gaze-stabilizing central vestibular neurons project asymmetrically to extraocular motoneuron pools
Zebrafish tilt-sensors and eye-movement are connected asymmetrically, suggesting that oculomotor input incorporates both the behavior of the species as well as evolved reflexive behaviors. [ Read More ]
02 — DeepEM3D: approaching human-level performance on 3D anisotropic EM image segmentation
Segmenting neuroimagery is hard. Determining airtight neurite boundaries makes the segmentation process way easier. [ Read More ]
01 — How many dinosaur species were there?
We can extrapolate how many dinosaur species there were using what we know about sampling bias, geological conditions, and the data from actual found fossils. [ Read More ]
31 — Fake News in Social Networks
The spread of fake news can be modeled as a graph where each node is a recurrent neural network that makes an opinionated decision of validity. [ Read More ]
30 — Synthesizing Deep Neural Network Architectures using Biological Synaptic Strength Distributions
Using synapse-strength distributions taken from biological cortex can improve computational neural net performance. [ Read More ]
29 — Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates
Using clusterwise inference in fMRI studies has likely led to pervasive, industry-wide, accidental inflation of statistical significance. [ Read More ]
28 — Underwater components of humpback whale bubble-net feeding behaviour
For the first time, individual whale bubble-blowing is correlated with 3D underwater location to characterize a complex feeding behavior. [ Read More ]
27 — Cofields: a physically inspired approach to motion coordination
Modeling a navigation-space as a physical "field" (e.g. magnetic) greatly simplifies the computational overhead for individual navigators in that space. [ Read More ]
26 — Introducing MantisBot: Hexapod robot controlled by a high-fidelity, real-time neural simulation
The authors demonstrate complex pose-maintenance and weight-distribution behaviors in a robot, using joint-positioning logic from insect neurons. [ Read More ]
25 — Electron Microscopic Reconstruction of Functionally Identified Cells in a Neural Integrator
Overlaying functional data with anatomical (structural) data helps identify and reconstruct complex brain circuits. [ Read More ]
24 — Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
By training cyclically adversarial neural networks simultaneously, it's possible to develop a nuanced image-net with limited or no training data. [ Read More ]
23 — Inferring cortical function in the mouse visual system through large-scale systems neuroscience.
Large brain circuits can be simulated at-scale by modeling neural systems as populations to reduce computational overhead and improve biofidelity. [ Read More ]
22 — Brain Computation Is Organized via Power-of-Two-Based Permutation Logic
One compelling hypothesis about how sensory signals are processed in the brain is substantiated by measuring from real brain tissue. [ Read More ]