# idealo/image-quality-assessment

Convolutional Neural Networks to predict the aesthetic and technical quality of images.

Repository: https://github.com/idealo/image-quality-assessment
Canonical: https://ross.abutalabs.com/products/image-quality-assessment
Homepage: https://idealo.github.io/image-quality-assessment/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: nima, neural-network, tensorflow, keras, mobilenet, aws, image-quality-assessment, convolutional-neural-networks, deep-learning, computer-vision, machine-learning, e-commerce, idealo
Archived: true
Last push: 2024-07-12T19:16:31+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3004, "days_push": 782, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2243, forks 455 (observed 2026-08-28T04:06:29.904206+00:00)

## What it is
A Python implementation of Google's NIMA (Neural Image Assessment) models that predict the aesthetic and technical quality of images using fine-tuned CNNs like MobileNet. It ships pre-trained models on the AVA and TID2013 datasets plus Docker images for CPU training locally and GPU training on AWS EC2.

## Use cases
- score the aesthetic quality of photos automatically
- rank millions of product or hotel images by visual appeal
- predict technical image quality like noise and blur
- filter low-quality user-uploaded images
- train an image quality model on custom data
- run NIMA image assessment in a Docker container

## When to choose
- you need pre-trained aesthetic or technical image quality scoring with Keras/TensorFlow
- you want to fine-tune NIMA models on your own image dataset
- you need to rank large image collections by quality at scale

## When to avoid
- you need a maintained library with recent framework support (Python 3.6 era, sparse recent activity)
- you need general-purpose image processing rather than quality scoring
- you work outside TensorFlow/Keras ecosystems

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing, computer-vision
- domain: computer-vision, image-processing, machine-learning, deep-learning
- platform: python, cloud
- tags: nima, image-quality, aesthetics, tensorflow, keras, mobilenet, transfer-learning, aws, docker, gpu

## Member repositories
- idealo/image-quality-assessment (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:29.904206+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 2026-08-30T02:44:03.623597+00:00, confidence not recorded.
  - readme: https://github.com/idealo/image-quality-assessment (fetched 2026-08-28T04:06:29.904206+00:00, sha c14268497376)
  - homepage: https://idealo.github.io/image-quality-assessment/ (fetched 2026-08-29T10:24:24.582569+00:00, sha b4e2bf2f0398)
- Data as of 2026-08-30T08:39:29.467469+00:00.