Papers › SOLQ: Segmenting Objects by Learning Queries

SOLQ: Segmenting Objects by Learning Queries

4 Jun 2021NeurIPS 2021 12arXiv:2106.02351archive 2025-07-28

Bin Dong, Fangao Zeng, Tiancai Wang, Xiangyu Zhang, Yichen Wei

In this paper, we propose an end-to-end framework for instance segmentation. Based on the recently introduced DETR [1], our method, termed SOLQ, segments objects by learning unified queries. In SOLQ, each query represents one object and has multiple representations: class, location and mask. The object queries learned perform classification, box regression and mask encoding simultaneously in an unified vector form. During training phase, the mask vectors encoded are supervised by the compression coding of raw spatial masks. In inference time, mask vectors produced can be directly transformed to spatial masks by the inverse process of compression coding. Experimental results show that SOLQ can achieve state-of-the-art performance, surpassing most of existing approaches. Moreover, the joint learning of unified query representation can greatly improve the detection performance of DETR. We hope our SOLQ can serve as a strong baseline for the Transformer-based instance segmentation. Code is available at https://github.com/megvii-research/SOLQ.

PaperPDFConference PDFCode

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

Code

megvii-research/SOLQ officialmentioned in papermentioned on GitHubpytorchNOASSERTION 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 Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Instance Segmentation COCO test-dev SOLQ (Swin-L, single scale) mask AP 46.7 #35 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev SOLQ (ResNet101, single scale) mask AP 40.9 #65 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev SOLQ (ResNet50, single scale) mask AP 39.7 #76 of 112 Archive leaderboard report
Object Detection COCO minival SOLQ (Swin-L, single scale) AP50 74.9 #212 of 220 Archive leaderboard report
Object Detection COCO minival SOLQ (Swin-L, single scale) AP75 61.3 #212 of 220 Archive leaderboard report
Object Detection COCO minival SOLQ (Swin-L, single scale) APL 71.9 #212 of 220 Archive leaderboard report
Object Detection COCO test-dev SOLQ (Swin-L, single scale) AP50 74.6 #41 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (Swin-L, single scale) AP75 60.5 #41 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (Swin-L, single scale) APL 70.6 #41 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (Swin-L, single scale) APM 60 #41 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (Swin-L, single scale) APS 37.6 #41 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (Swin-L, single scale) box mAP 56.5 #41 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (ResNet101, single scale) box mAP 48.7 #103 of 225 Archive leaderboard report
Object Detection COCO test-dev SOLQ (ResNet50, single scale) box mAP 47.8 #115 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

ConvolutionDense ConnectionsDetrFeedforward NetworkSoftmax

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