facebookresearch/detr

Health v2 · maintenance only

Flags: archived

Adoption not part of the score

15354 stars · 2669 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

DETR is Facebook Research's PyTorch implementation of Detection Transformer, an end-to-end object detection model that replaces hand-crafted pipelines with a Transformer encoder-decoder and bipartite matching loss. It ships pretrained models matching Faster R-CNN on COCO with simple training and inference code.

Use cases

• detect objects in images with a transformer model
• train an end-to-end object detector on custom data
• run pretrained COCO object detection inference in a few lines of PyTorch
• compare DETR with Faster R-CNN baselines
• fine-tune a detection transformer on my dataset

When to choose

• you want a simple, research-friendly object detection codebase without complex detection libraries
• you need transformer-based set prediction detection with pretrained COCO weights

When to avoid

• you need production-grade detection tooling with extensive model zoo and deployment support
• you need the latest detection architectures or frequent updates

Facets

library · maturity maintenance

computer-vision machine-learning deep-learning computer-vision deep-learning machine-learning python object-detection transformer pytorch set-prediction coco research gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="facebookresearch/detr")