A latent text-to-image diffusion model

CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image

The image prompt adapter is designed to enable a pretrained text-to-image diffusion model to generate images with image prompt.

Code repo for realtime multi-person pose estimation in CVPR'17 (Oral)

Acceptance rates for the major AI conferences

Simplified implementations of deep learning related works

PyTorch implementation of SimCLR: A Simple Framework for Contrastive Learning of Visual Representations

pytorch implementation of openpose including Hand and Body Pose Estimation.

[CVPR 2021] Official PyTorch implementation for Transformer Interpretability Beyond Attention Visualization, a novel method to visualize classifications by Transformer based networks.

Efficient computing methods developed by Huawei Noah's Ark Lab

ICCV2021, Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet

Code for ICLR 2020 paper "VL-BERT: Pre-training of Generic Visual-Linguistic Representations".

Multimodal Models in Real World

Person Image Synthesis via Denoising Diffusion Model (CVPR 2023)

[SIGGRAPH Asia 2022] IDE-3D: Interactive Disentangled Editing For High-Resolution 3D-aware Portrait Synthesis

Epipolar Transformers (best paper award, CVPR 2020 workshop)

PyTorch Implementation for "TransPose: Keypoint localization via Transformer", ICCV 2021.

Code repo for "LSTM Pose Machines" (CVPR'18)

MoVQGAN - model for the image encoding and reconstruction

(CVPR 2021) PRTR: Pose Recognition with Cascade Transformers

[WACV 2020] "Nonparametric Structure Regularization Machine for 2D Hand Pose Estimation"

CVPR 2024 "Instance Tracking in 3D Scenes from Egocentric Videos"

Pytorch implementation on PennAction (LSTM_Pose_Machines_CVPR_2018_paper)