Setup and customize deep learning environment in seconds.

Pretrained EfficientNet, EfficientNet-Lite, MixNet, MobileNetV3 / V2, MNASNet A1 and B1, FBNet, Single-Path NAS

PyTorch implementation of the YOLO (You Only Look Once) v2

Caffe2 implementation of Open Neural Network Exchange (ONNX)

Deep Learning Benchmarking Suite

PyTorch implementation of the OpenPose

Convert image folder to lmdb, adapted from Efficient-PyTorch

CNN model inference benchmarks for some popular deep learning frameworks

Domain Adaptive Faster R-CNN in Detectron

Caffe2 C++/Python tutorial with full demo

Tensorboard for Caffe2

Covert original YOLO model from Pytorch to Onnx, and do inference using backend Caffe2 or Tensorflow.

PiNN2 is a easy-to-use framework for device compact modeling using physics-inspired neural networks

End-to-end multitask CNN architecture for object instance segmentation, human pose detection and multi-person tracking based on Facebook AI's Detectron system

A real time face mask detection model

Device Agnostic Deep Learning Benchmark to compare the performance framework-wise, layer-wise, and model-wise.

Contains some of my experiments using caffe2

This repository shows an example of how to use the ONNX standard to interoperate between different frameworks. In this example, we train a model with PyTorch and make predictions with Tensorflow, ONNX Runtime, and Caffe2.

Tattoo detection and localization

Caffe2-vision help engineer to build vision CNN for training and product, including dataset maker and favorite models