Mask DINO
One transformer that does detection, instance, panoptic, and semantic segmentation with a single set of weights.
About
Mask DINO extends the DINO detector with a mask prediction branch, unifying object detection and all three image segmentation tasks (instance, panoptic, semantic) in one architecture and one training process. Query embeddings from the DINO decoder are dot-producted against a high-resolution pixel embedding map to produce binary masks, and the shared backbone lets detection and segmentation data cooperate during training instead of competing.
The paper's headline numbers, all with models under one billion parameters: 54.5 AP on COCO instance segmentation, 59.4 PQ on COCO panoptic, and 60.8 mIoU on ADE20K semantic segmentation, each the best reported at the time of publication. It is Apache-2.0 licensed, which makes it one of the most commercially friendly high-accuracy segmentation frameworks in the catalog.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Author-reported: 54.5 AP COCO instance, 59.4 PQ COCO panoptic, 60.8 mIoU ADE20K (models under 1B params)
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: paper (arXiv:2206.02777, CVPR 2023) with COCO/PQ/mIoU numbers confirmed via the abstract and the official repo; Apache-2.0 confirmed via downstream attribution. Strengths: unified detection plus all segmentation tasks, SOTA numbers at release, permissive license. Docked for: (1) Detectron2-based codebase is heavy to set up; (2) repo activity has slowed as the authors moved to detrex. License: Apache-2.0.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 6.0/10
- Openness · license + docs
- 9.0/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · CVPR 2023 (arXiv 2206.02777, Jun 2022)
- · 54.5 AP COCO instance / 59.4 PQ COCO panoptic / 60.8 mIoU ADE20K, author-reported
- · Unified detection + instance/panoptic/semantic segmentation