LoSealL/VideoSuperResolution

Health v2 · maintenance only

Adoption not part of the score

1687 stars · 295 forks observed · 2026-08-28

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

A Python library (pip-installable as VSR) collecting reimplementation of state-of-the-art single-image and video super-resolution neural network architectures in TensorFlow, with some PyTorch ports. It includes many classic models such as SRCNN, VDSR, EDSR, SRGAN, and DBPN, some with downloadable pretrained weights.

Use cases

• upscale low-resolution videos with deep learning
• enhance resolution of single images using SR models
• compare super-resolution architectures like EDSR and SRGAN
• reproduce benchmark results from NTIRE super-resolution papers
• run pretrained super-resolution models in TensorFlow
• experiment with video super-resolution networks like VESPCN

When to choose

• you want many SR architectures implemented in one consistent framework
• you need a pip-installable super-resolution library with pretrained weights
• you are researching or benchmarking image/video upscaling models

When to avoid

• you need production-grade, actively maintained upscaling software
• you require the latest 2021+ architectures or GPU-optimized inference
• you work outside Python/TensorFlow ecosystems

Facets

library · maturity maintenance

machine-learning image-processing video-processing deep-learning computer-vision image-processing deep-learning machine-learning python cross-platform super-resolution tensorflow pytorch srgan edsr vdsr video-upscaling image-upscaling pretrained-models video

Member repositories

For agents

markdown · JSON · MCP: product_card(name="LoSealL/VideoSuperResolution")