Papers › ISTR: End-to-End Instance Segmentation with Transformers

ISTR: End-to-End Instance Segmentation with Transformers

3 May 2021arXiv:2105.00637archive 2025-07-28

Jie Hu, Liujuan Cao, Yao Lu, Shengchuan Zhang, Yan Wang, Ke Li, Feiyue Huang, Ling Shao, Rongrong Ji

End-to-end paradigms significantly improve the accuracy of various deep-learning-based computer vision models. To this end, tasks like object detection have been upgraded by replacing non-end-to-end components, such as removing non-maximum suppression by training with a set loss based on bipartite matching. However, such an upgrade is not applicable to instance segmentation, due to its significantly higher output dimensions compared to object detection. In this paper, we propose an instance segmentation Transformer, termed ISTR, which is the first end-to-end framework of its kind. ISTR predicts low-dimensional mask embeddings, and matches them with ground truth mask embeddings for the set loss. Besides, ISTR concurrently conducts detection and segmentation with a recurrent refinement strategy, which provides a new way to achieve instance segmentation compared to the existing top-down and bottom-up frameworks. Benefiting from the proposed end-to-end mechanism, ISTR demonstrates state-of-the-art performance even with approximation-based suboptimal embeddings. Specifically, ISTR obtains a 46.8/38.6 box/mask AP using ResNet50-FPN, and a 48.1/39.9 box/mask AP using ResNet101-FPN, on the MS COCO dataset. Quantitative and qualitative results reveal the promising potential of ISTR as a solid baseline for instance-level recognition. Code has been made available at: https://github.com/hujiecpp/ISTR.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

hujiecpp/ISTR officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Instance SegmentationObject DetectionSegmentationSemantic Segmentationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Instance Segmentation COCO test-dev ISTR-SMT (Swin-L, single scale) mask AP 49.7 #23 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) APL 52.3 #73 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) APM 41.9 #73 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) APS 22.8 #73 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) mask AP 39.9% #73 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) APL 50.6 #85 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) APM 40.4 #85 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) APS 22.1 #85 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) mask AP 38.6% #85 of 112 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) APL 59.9 #43 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) APM 48.7 #43 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) APS 27.8 #43 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet50-FPN-3x, single-scale) box mAP 56.4 #43 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) APL 61.5 #110 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) APM 50.4 #110 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) APS 28.7 #110 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet101-FPN-3x, single-scale) box mAP 48.1 #110 of 225 Archive leaderboard report
Object Detection COCO test-dev ISTR (ResNet50-FPN-3x) box mAP 46.8 #123 of 225 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections