A flexible framework of neural networks for deep learning

A simplified implemention of Faster R-CNN that replicate performance from origin paper

NumPy & SciPy for GPU

Flexible and powerful tensor operations for readable and reliable code (for pytorch, jax, TF and others)

ChainerCV: a Library for Deep Learning in Computer Vision

TensorLy: Tensor Learning in Python.

an implementation of 3D Ken Burns Effect from a Single Image using PyTorch

an implementation of Video Frame Interpolation via Adaptive Separable Convolution using PyTorch

a reimplementation of PWC-Net in PyTorch that matches the official Caffe version

a reimplementation of LiteFlowNet in PyTorch that matches the official Caffe version

an implementation of softmax splatting for differentiable forward warping using PyTorch

Chainer implementation of waifu2x

Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset (CVPR'19)

Library for faster pinned CPU <-> GPU transfer in Pytorch

CuPy fused PyTorch neural networks ops

Deep learning framework realized by Numpy purely, supports for both Dynamic Graph and Static Graph with GPU acceleration

The code for multi-channel source separation and dereverberation such as FastMNMF1, FastMNMF2, and AR-FastMNMF2.

Official source code for our paper "AdaCoF: Adaptive Collaboration of Flows for Video Frame Interpolation" (CVPR 2020)

Computer vision container that includes Jupyter notebooks with built-in code hinting, Anaconda, CUDA 11.8, TensorRT inference accelerator for Tensor cores, CuPy (GPU drop in replacement for Numpy), PyTorch, PyTorch geometric for Graph Neural Networks, TF2, Tensorboard, and OpenCV for accelerated workloads on NVIDIA Tensor cores and GPUs.

A suite of benchmarks for CPU and GPU performance of the most popular high-performance libraries for Python 🚀