Okazawa lab's starred repositories
pytorch-image-models
The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more
robustness
A library for experimenting with, training and evaluating neural networks, with a focus on adversarial robustness.
BigGAN-PyTorch
The author's officially unofficial PyTorch BigGAN implementation.
opticka
Opticka is an experiment manager built on top of the Psychophysics toolbox (PTB) for MATLAB. It runs experimental tasks using flexible state machine logic and easily does dynamic methods-of-constants type experiments with full behavioural control. It uses a class system to create simple to use visual stimuli using experimenter friendly units. Opticka can use an Eyelink eyetracker, Plexon Omniplex neurophysiology data collection and general TTL control using either a cheap LabJack (ms precision) or DataPixx/Display++ (µs precision). It contains analysis routines linked to Fieldtrip for spike and LFP data easily parsed in terms of the experimental variables.