PanSR: An Object-Centric Mask Transformer for Panoptic Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Žust, Lojze, Kristan, Matej
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910745107628032
author Žust, Lojze
Kristan, Matej
author_facet Žust, Lojze
Kristan, Matej
contents Panoptic segmentation is a fundamental task in computer vision and a crucial component for perception in autonomous vehicles. Recent mask-transformer-based methods achieve impressive performance on standard benchmarks but face significant challenges with small objects, crowded scenes and scenes exhibiting a wide range of object scales. We identify several fundamental shortcomings of the current approaches: (i) the query proposal generation process is biased towards larger objects, resulting in missed smaller objects, (ii) initially well-localized queries may drift to other objects, resulting in missed detections, (iii) spatially well-separated instances may be merged into a single mask causing inconsistent and false scene interpretations. To address these issues, we rethink the individual components of the network and its supervision, and propose a novel method for panoptic segmentation PanSR. PanSR effectively mitigates instance merging, enhances small-object detection and increases performance in crowded scenes, delivering a notable +3.4 PQ improvement over state-of-the-art on the challenging LaRS benchmark, while reaching state-of-the-art performance on Cityscapes. The code and models will be publicly available at https://github.com/lojzezust/PanSR.
format Preprint
id arxiv_https___arxiv_org_abs_2412_10589
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PanSR: An Object-Centric Mask Transformer for Panoptic Segmentation
Žust, Lojze
Kristan, Matej
Computer Vision and Pattern Recognition
Panoptic segmentation is a fundamental task in computer vision and a crucial component for perception in autonomous vehicles. Recent mask-transformer-based methods achieve impressive performance on standard benchmarks but face significant challenges with small objects, crowded scenes and scenes exhibiting a wide range of object scales. We identify several fundamental shortcomings of the current approaches: (i) the query proposal generation process is biased towards larger objects, resulting in missed smaller objects, (ii) initially well-localized queries may drift to other objects, resulting in missed detections, (iii) spatially well-separated instances may be merged into a single mask causing inconsistent and false scene interpretations. To address these issues, we rethink the individual components of the network and its supervision, and propose a novel method for panoptic segmentation PanSR. PanSR effectively mitigates instance merging, enhances small-object detection and increases performance in crowded scenes, delivering a notable +3.4 PQ improvement over state-of-the-art on the challenging LaRS benchmark, while reaching state-of-the-art performance on Cityscapes. The code and models will be publicly available at https://github.com/lojzezust/PanSR.
title PanSR: An Object-Centric Mask Transformer for Panoptic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.10589