# OpenGVLab/InternImage

[CVPR 2023 Highlight] InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions

Repository: https://github.com/OpenGVLab/InternImage
Canonical: https://ross.abutalabs.com/products/internimage
Homepage: https://arxiv.org/abs/2211.05778
Language: Python
License: MIT
License Family: permissive
Topics: backbone, deformable-convolution, object-detection, semantic-segmentation, foundation-model
Last push: 2025-03-25T02:29:19+00:00

## Health v2 (maintenance only)
Score: 28/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 13, release rhythm 8, longevity 99
- inputs: {"age_days": 1392, "days_push": 527, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2841, forks 265 (observed 2026-08-28T04:07:24.791156+00:00)

## What it is
InternImage is a large-scale CNN-based vision foundation model that uses deformable convolutions as its core operator, released with pretrained weights and code. It achieves state-of-the-art results on ImageNet classification, COCO object detection, and ADE20K semantic segmentation.

## Use cases
- pretrained backbone for object detection
- semantic segmentation model for ADE20K or Cityscapes
- image classification with a large-scale CNN
- fine-tune a vision foundation model on custom detection data
- deformable convolution implementation in PyTorch
- state-of-the-art COCO detection model

## When to choose
- you need a strong pretrained vision backbone for detection or segmentation
- you want CNN-based alternatives to ViT foundation models
- you need top accuracy on COCO or ADE20K benchmarks

## When to avoid
- you need lightweight models for edge or mobile deployment
- you lack GPU resources for large-scale training or inference
- you need a general-purpose ML framework rather than a specific model family

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, deep-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning, image-processing
- platform: python
- tags: vision-backbone, deformable-convolution, object-detection, semantic-segmentation, foundation-model, cnn, pretrained-models, gpu, linux

## Member repositories
- OpenGVLab/InternImage (main) score 28

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:24.791156+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 2026-08-30T07:37:14.712432+00:00, confidence not recorded.
  - readme: https://github.com/OpenGVLab/InternImage (fetched 2026-08-28T04:07:24.791156+00:00, sha e45c8cf6ade1)
  - homepage: https://arxiv.org/abs/2211.05778 (fetched 2026-08-29T09:52:53.496799+00:00, sha 0cffc3cdb057)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T09:52:53.499266+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T09:52:53.502626+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T09:52:53.504557+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T09:52:53.500931+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.