ckkelvinchan/RealBasicVSR

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1100 stars · 150 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Official PyTorch implementation of RealBasicVSR, a real-world video super-resolution model from CVPR 2022, built on MMEditing. It provides training/inference code, pre-trained weights, a Colab demo, and the VideoLQ benchmark dataset.

Use cases

• upscale low-quality real-world videos
• video super-resolution with pretrained model
• denoise and enhance compressed video footage
• benchmark video restoration on VideoLQ dataset
• train a video super-resolution model
• restore old or blurry video clips

When to choose

• you need state-of-the-art real-world video super-resolution with pretrained weights
• you want a research-grade codebase built on MMEditing for video restoration
• you need the VideoLQ benchmark for evaluating video restoration models

When to avoid

• you need real-time or low-latency upscaling - the model is computationally heavy
• you want a maintained production library - development has largely moved to MMEditing
• you only need simple image (not video) super-resolution

Facets

library · maturity maintenance

video-processing image-processing machine-learning deep-learning computer-vision image-processing deep-learning python super-resolution video-super-resolution cvpr-2022 pytorch mmediting research-code pretrained-models video gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="ckkelvinchan/RealBasicVSR")