Papers › RDSNet: A New Deep Architecture for Reciprocal Object Detection and Instance Segmentation

RDSNet: A New Deep Architecture for Reciprocal Object Detection and Instance Segmentation

11 Dec 2019arXiv:1912.05070archive 2025-07-28

Shaoru Wang, Yongchao Gong, Junliang Xing, Lichao Huang, Chang Huang, Weiming Hu

Object detection and instance segmentation are two fundamental computer vision tasks. They are closely correlated but their relations have not yet been fully explored in most previous work. This paper presents RDSNet, a novel deep architecture for reciprocal object detection and instance segmentation. To reciprocate these two tasks, we design a two-stream structure to learn features on both the object level (i.e., bounding boxes) and the pixel level (i.e., instance masks) jointly. Within this structure, information from the two streams is fused alternately, namely information on the object level introduces the awareness of instance and translation variance to the pixel level, and information on the pixel level refines the localization accuracy of objects on the object level in return. Specifically, a correlation module and a cropping module are proposed to yield instance masks, as well as a mask based boundary refinement module for more accurate bounding boxes. Extensive experimental analyses and comparisons on the COCO dataset demonstrate the effectiveness and efficiency of RDSNet. The source code is available at https://github.com/wangsr126/RDSNet.

PaperPDFCode

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

Code

wangsr126/RDSNet officialmentioned in papermentioned on GitHubpytorchApache-2.0 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 SegmentationObjectObject DetectionSegmentationSemantic SegmentationTranslationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Instance Segmentation COCO test-dev RDSNet (data aug) AP50 57.9% #96 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev RDSNet (data aug) AP75 39.0% #96 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev RDSNet (data aug) APL 51.6% #96 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev RDSNet (data aug) APM 39.5% #96 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev RDSNet (data aug) APS 16.4% #96 of 112 Archive leaderboard report
Instance Segmentation COCO test-dev RDSNet (data aug) mask AP 36.4% #96 of 112 Archive leaderboard report
Object Detection COCO test-dev RDSNet (ResNet-101, RetinaNet, mask, MBRM) AP50 60.1 #195 of 225 Archive leaderboard report
Object Detection COCO test-dev RDSNet (ResNet-101, RetinaNet, mask, MBRM) AP75 43 #195 of 225 Archive leaderboard report
Object Detection COCO test-dev RDSNet (ResNet-101, RetinaNet, mask, MBRM) APL 51.5 #195 of 225 Archive leaderboard report
Object Detection COCO test-dev RDSNet (ResNet-101, RetinaNet, mask, MBRM) APM 43.5 #195 of 225 Archive leaderboard report
Object Detection COCO test-dev RDSNet (ResNet-101, RetinaNet, mask, MBRM) APS 22.1 #195 of 225 Archive leaderboard report
Object Detection COCO test-dev RDSNet (ResNet-101, RetinaNet, mask, MBRM) box mAP 40.3 #195 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

1x1 ConvolutionAverage PoolingBatch NormalizationBottleneck Residual BlockConvolutionGlobal Average PoolingKaiming InitializationMax PoolingReLUResidual BlockResidual Connection

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