1. EnhanceGAN Torch

    An adversarial learning based model that performs automatic image enhancement. It only requires weak supervision (binary labels on image aesthetic quality) and is able to learn enhancement operators for the task of aesthetic-based image enhancement
    Aesthetic-Driven Image Enhancement by Adversarial Learning, ACM MM 2018

  2. LiteFlowNet Caffe

    A light-weight convolutional neural network for optical flow estimation. Over 30 times smaller in model size, and 1.36 times faster in speed than FlowNet2
    LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow Estimation, CVPR 2018

  3. RL-Restore TensorFlow

    Image restoration by reinforcement learning
    Crafting a Toolchain for Image Restoration by Deep Reinforcement Learning, CVPR 2018

  4. SFT-GAN Torch, PyTorch

    Single image super-resolution with image segmentation prior
    Recovering Realistic Texture in Image Super-resolution by Deep Spatial Feature Transform, CVPR 2018

  5. BasicSR PyTorch

    An image super-resolution toolkit that covers popular methods such as SRCNN, SRResNet, SRGAN, etc.

  6. DREAM PyTorch

    A simple block that can be used to perform profile-to-frontal face transformation in the deep feature space
    Pose-Robust Face Recognition via Deep Residual Equivariant Mapping, CVPR 2018

  7. Mix-and-Match Caffe

    A network tuning approach that allows image segmentation networks pre-trained with self-supervised learning to outperform the supervised learning counterpart
    Mix-and-Match Tuning for Self-Supervised Semantic Segmentation, AAAI 2018

  8. FashionGAN Torch

    Generating new clothing on a wearer through generative adversarial learning
    Be Your Own Prada: Fashion Synthesis with Structural Coherence, ICCV 2017

  9. LMLE Caffe

    Large Margin Local Embedding, a method for learning deep represenatation from data with imbalanced class distribution through quintuplet sampling and triple-header hinge loss
    Learning Deep Representation for Imbalanced Classification, CVPR 2016

  10. Slicing-CNN Caffe

    An end-to-end model named as Slicing CNN (S-CNN) for crowd video understanding
    Slicing Convolutional Neural Network for Crowd Video Understanding, CVPR 2016


    An initialization-free face alignment method that is robust to large pose variation
    Face Alignment by Coarse-to-Fine Shape Searching, ICCV 2015

  12. CBN Caffe

    A deep convolutional neural network for human face hallucination
    Deep Cascaded Bi-Network for Face Hallucination, ECCV 2016

  13. MSG-Net Caffe

    A deep convolutional neural network for depth map image super-resolution
    Depth Map Super Resolution by Deep Multi-Scale Guidance, ECCV 2016

  14. FSRCNN Caffe

    A convolutional neural network for image super-resolution, faster than SRCNN
    Accelerating the Super-Resolution Convolutional Neural Network, ECCV 2016

  15. ARCNN Caffe

    A convolutional neural network for reducing artifacts caused by image compression
    Compression Artifacts Reduction by a Deep Convolutional Network, ICCV 2015

  16. WWW Deep Crowd Caffe

    A two-branch CNN model (i.e. appearance branch and motion branch) for predicting 94 crowd attributes
    Deeply Learned Attributes for Crowded Scene Understanding, CVPR 2015

  17. SRCNN Caffe

    The first convolutional neural network for single image super-resolution
    Learning a Deep Convolutional Network for Image Super-Resolution, ECCV 2014


    A deep convolutional neural network that simultaneously predicts facial landmarks and facial attributes
    Facial Landmark Detection by Deep Multi-task Learning, ECCV 2014

  19. Crowd Profiling MATLAB

    Source code for scene independent group profiling in crowd
    Scene-Independent Group Profiling in Crowd, CVPR 2014