# IDEA-Research/detrex

detrex is a research platform for DETR-based object detection, segmentation, pose estimation and other visual recognition tasks.

Repository: https://github.com/IDEA-Research/detrex
Canonical: https://ross.abutalabs.com/products/detrex
Homepage: https://detrex.readthedocs.io/en/latest/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: detr, object-detection, pytorch, dino, state-of-the-art, dab-detr, deformable-detr, conditional-detr, dn-detr, group-detr, h-detr, mask-dino, anchor-detr, deta, pose-estimation, segmentation
Last push: 2025-09-11T12:21:01+00:00

## Health v2 (maintenance only)
Score: 41/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 41, release rhythm 8, longevity 100
- inputs: {"age_days": 1493, "days_push": 356, "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 2306, forks 248 (observed 2026-08-28T04:06:35.771177+00:00)

## What it is
detrex is an open-source PyTorch-based research platform and toolbox for DETR-style Transformer detection algorithms, built on top of Detectron2. It provides modular implementations of state-of-the-art models like DINO, Deformable-DETR, and Mask DINO for object detection, segmentation, and pose estimation.

## Use cases
- train a DETR-based object detection model
- benchmark detection transformer algorithms
- run instance segmentation with Mask DINO
- reproduce DINO or Deformable-DETR results
- build custom transformer detection architectures
- do pose estimation with transformer models

## When to choose
- you need state-of-the-art DETR-family detection models with pretrained weights
- you want a modular Detectron2-based platform for detection research
- you're benchmarking transformer-based detection, segmentation, or pose estimation methods

## When to avoid
- you need production-ready detection rather than a research codebase
- you prefer lightweight CNN detectors like YOLO or Faster R-CNN
- you don't work with PyTorch or GPUs

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning
- platform: python
- tags: detr, object-detection, transformer, pytorch, segmentation, pose-estimation, detectron2, model-zoo, research-toolbox, gpu, linux

## Member repositories
- IDEA-Research/detrex (main) score 41

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:35.771177+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 2026-08-30T02:39:40.698217+00:00, confidence not recorded.
  - readme: https://github.com/IDEA-Research/detrex (fetched 2026-08-28T04:06:35.771177+00:00, sha 331d9d2a3409)
- Data as of 2026-08-30T08:39:29.467469+00:00.