# xinntao/ESRGAN

ECCV18 Workshops - Enhanced SRGAN. Champion PIRM Challenge on Perceptual Super-Resolution. The training codes are in BasicSR.

Repository: https://github.com/xinntao/ESRGAN
Canonical: https://ross.abutalabs.com/products/esrgan
Homepage: https://github.com/xinntao/BasicSR
Language: Python
License: Apache-2.0
License Family: permissive
Topics: esrgan
Last push: 2022-10-19T02:33:55+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2924, "days_push": 1415, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6568, forks 1119 (observed 2026-08-28T04:09:45.291636+00:00)

## What it is
ESRGAN (Enhanced SRGAN) is a PyTorch-based image super-resolution model that won the PIRM 2018 Challenge on Perceptual Super-Resolution. This repository provides testing code, pretrained models, and a network interpolation demo, with training code moved to BasicSR and development continued in Real-ESRGAN.

## Use cases
- upscale low-resolution images with better perceptual quality
- enhance old photos by increasing resolution
- test pretrained super-resolution models in python
- interpolate between SRGAN and ESRGAN models
- research perceptual image super-resolution

## When to choose
- you need a proven perceptual super-resolution model with pretrained weights
- you want a simple inference/testing setup for ESRGAN
- you are reproducing ECCV18 workshop or PIRM challenge results

## When to avoid
- you need training code - use BasicSR instead
- you want practical real-world restoration with artifact removal - use Real-ESRGAN
- you need tile inference, alpha channel, or 16-bit image support - available in Real-ESRGAN
- you need an actively developed tool - this repo is frozen for compatibility

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: computer-vision, image-processing, deep-learning
- platform: python, windows
- tags: super-resolution, esrgan, srgan, image-restoration, pytorch, pretrained-models, pirm-challenge, linux, macos

## Member repositories
- xinntao/ESRGAN (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:45.291636+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:43:39.795263+00:00, confidence not recorded.
  - readme: https://github.com/xinntao/ESRGAN (fetched 2026-08-28T04:09:45.291636+00:00, sha 59b2d5a8ec81)
  - homepage: https://github.com/xinntao/BasicSR (fetched 2026-08-29T08:40:15.039592+00:00, sha 5d2ec73c7496)
- Data as of 2026-08-30T08:39:29.467469+00:00.