Papers › Multi-task Fusion for Efficient Panoptic-Part Segmentation

Multi-task Fusion for Efficient Panoptic-Part Segmentation

15 Dec 2022arXiv:2212.07671archive 2025-07-28

Sravan Kumar Jagadeesh, René Schuster, Didier Stricker

In this paper, we introduce a novel network that generates semantic, instance, and part segmentation using a shared encoder and effectively fuses them to achieve panoptic-part segmentation. Unifying these three segmentation problems allows for mutually improved and consistent representation learning. To fuse the predictions of all three heads efficiently, we introduce a parameter-free joint fusion module that dynamically balances the logits and fuses them to create panoptic-part segmentation. Our method is evaluated on the Cityscapes Panoptic Parts (CPP) and Pascal Panoptic Parts (PPP) datasets. For CPP, the PartPQ of our proposed model with joint fusion surpasses the previous state-of-the-art by 1.6 and 4.7 percentage points for all areas and segments with parts, respectively. On PPP, our joint fusion outperforms a model using the previous top-down merging strategy by 3.3 percentage points in PartPQ and 10.5 percentage points in PartPQ for partitionable classes.

PaperPDF

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

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

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

Tasks

Image SegmentationPart-aware Panoptic SegmentationRepresentation LearningSegmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Segmentation Pascal Panoptic Parts JPPF mIoUPartS 54.4 #4 of 4 Archive leaderboard report
Part-aware Panoptic Segmentation Cityscapes Panoptic Parts JPPF PartPQ 61.8 #4 of 4 Archive leaderboard report
Part-aware Panoptic Segmentation Pascal Panoptic Parts JPPF PartPQ 32.3 #4 of 4 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.

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