Skip to content

Latest commit

 

History

History
66 lines (54 loc) · 3.63 KB

README.md

File metadata and controls

66 lines (54 loc) · 3.63 KB
Raw

Enhancing Perceptual Quality in Video Super-Resolution through Temporally-Consistent Detail Synthesis using Diffusion Models

Accepted to ECCV 2024

Paper

Abstract

In this paper, we address the problem of enhancing perceptual quality in video super-resolution (VSR) using Diffusion Models (DMs) while ensuring temporal consistency among frames. We present StableVSR, a VSR method based on DMs that can significantly enhance the perceptual quality of upscaled videos by synthesizing realistic and temporally-consistent details. We introduce the Temporal Conditioning Module (TCM) into a pre-trained DM for single image super-resolution to turn it into a VSR method. TCM uses the novel Temporal Texture Guidance, which provides it with spatially-aligned and detail-rich texture information synthesized in adjacent frames. This guides the generative process of the current frame toward high-quality and temporally-consistent results. In addition, we introduce the novel Frame-wise Bidirectional Sampling strategy to encourage the use of information from past to future and vice-versa. This strategy improves the perceptual quality of the results and the temporal consistency across frames. We demonstrate the effectiveness of StableVSR in enhancing the perceptual quality of upscaled videos while achieving better temporal consistency compared to existing state-of-the-art methods for VSR.

Method overview

networkfull

Usage

Environment

The code is based on Python 3.8.17, CUDA 11, and diffusers.

Conda setup

conda create -n stablevsr python=3.8.17 -y
git clone https://github.com/claudiom4sir/StableVSR.git
cd StableVSR
conda activate stablevsr
pip install -r requirements.txt

Datasets

Download the REDS dataset from here (sharp + low-resolution). Data are expected to be in the format root/hr/sequences/frames and root/lr/sequences/frames.

Pretrained models

Pretrained models are available here. If you run the train or test code, you don't need to download them explicitly as they are fetched with .from_pretrained('claudiom4sir/StableVSR').

Train

Adjust the dataroot options in dataset/config_reds.yaml. Then, adjust the options in train.sh. Use the following command to start training:

bash ./train.sh

Test

python test.py --in_path YOUR_PATH_TO_LR_SEQS --out_path YOUR_OUTPUT_PATH --num_inference_steps 50 --controlnet_ckpt YOUR_PATH_TO_CONTROLNET_CKPT_FOLDER

Evaluation

python eval.py --gt_path YOUR_PATH_TO_GT_SEQS --out_path YOUR_OUTPUT_PATH

Memory requirements

Training with the provided configuration requires about 17GB GPU. Evaluation on REDS (320x180 -> 1280x720) about 14.5 GB.

Demo video

000.mp4
011.mp4
015.mp4
020.mp4

Citations

@article{rota2023enhancing,
  title={Enhancing Perceptual Quality in Video Super-Resolution through Temporally-Consistent Detail Synthesis using Diffusion Models},
  author={Rota, Claudio and Buzzelli, Marco and van de Weijer, Joost},
  journal={arXiv preprint arXiv:2311.15908},
  year={2023}
}

Contacts

If you have any questions, please contact me at claudio.rota@unimib.it