Papers › Mask-aware IoU for Anchor Assignment in Real-time Instance Segmentation

Mask-aware IoU for Anchor Assignment in Real-time Instance Segmentation

19 Oct 2021arXiv:2110.09734archive 2025-07-28

Kemal Oksuz, Baris Can Cam, Fehmi Kahraman, Zeynep Sonat Baltaci, Sinan Kalkan, Emre Akbas

This paper presents Mask-aware Intersection-over-Union (maIoU) for assigning anchor boxes as positives and negatives during training of instance segmentation methods. Unlike conventional IoU or its variants, which only considers the proximity of two boxes; maIoU consistently measures the proximity of an anchor box with not only a ground truth box but also its associated ground truth mask. Thus, additionally considering the mask, which, in fact, represents the shape of the object, maIoU enables a more accurate supervision during training. We present the effectiveness of maIoU on a state-of-the-art (SOTA) assigner, ATSS, by replacing IoU operation by our maIoU and training YOLACT, a SOTA real-time instance segmentation method. Using ATSS with maIoU consistently outperforms (i) ATSS with IoU by ∼1 mask AP, (ii) baseline YOLACT with fixed IoU threshold assigner by ∼2 mask AP over different image sizes and (iii) decreases the inference time by 25 % owing to using less anchors. Then, exploiting this efficiency, we devise maYOLACT, a faster and +6 AP more accurate detector than YOLACT. Our best model achieves $37.7$ mask AP at $25$ fps on COCO test-dev establishing a new state-of-the-art for real-time instance segmentation. Code is available at https://github.com/kemaloksuz/Mask-aware-IoU

PaperPDFCode

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

Code

kemaloksuz/mask-aware-iou 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 SegmentationReal-time Instance SegmentationSegmentationSemantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) AP50 59.4 #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) AP75 39.9 #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) APL 52.5 #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) APM 40.8 #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) APS 18.1 #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) Frame (fps) 25 (Tesla V100) #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-700 (ResNet-50) mask AP 37.7 #8 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) AP50 56.2 #12 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) AP75 37.1 #12 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) APL 51.4 #12 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) APM 38.0 #12 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) APS 14.7 #12 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) Frame (fps) 30 (Tesla V100) #12 of 22 Archive leaderboard report
Real-time Instance Segmentation MSCOCO maYOLACT-550 (ResNet-50) mask AP 35.2 #12 of 22 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

ATSS

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