Home

Awesome

BiFormer (CVPR2023)

Junheum Park, Jintae Kim, and Chang-Su Kim

Official code for "BiFormer: Learning Bilateral Motion Estimation via Bilateral Transformer for 4K Video Frame Interpolation"[paper]

Requirements

Installation

Create conda environment:

    $ conda create -n BiFormer python=3.9 anaconda
    $ conda activate BiFormer
    $ conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.6 -c pytorch -c conda-forge
    $ pip install opencv-python einops timm cupy-cuda116 easydict

Download repository:

    $ git clone https://github.com/JunHeum/BiFormer.git

Download pre-trained model parameters:

    $ unzip BiFormer_Weights.zip

Quick Usage

Generate an intermediate frame on your pair of frames:

    $ python run.py --first images/im1.png --second images/im3.png --output images/im2.png

Citation

Please cite the following paper if you feel this repository useful.

    @inproceedings{park2023BiFormer,
        author    = {Park, Junheum and Kim, Jintae and Kim, Chang-Su}, 
        title     = {BiFormer: Learning Bilateral Motion Estimation via Bilateral Transformer for 4K Video Frame Interpolation}, 
        booktitle = {Computer Vision and Pattern Recognition},
        year      = {2023}
    }

License

See Apache License