Advancing Complex Wide-Area Scene Understanding with Hierarchical Coresets Selection

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Jingyao, Chen, Yiming, Si, Lingyu, Zheng, Changwen
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918163523829760
author Wang, Jingyao
Chen, Yiming
Si, Lingyu
Zheng, Changwen
author_facet Wang, Jingyao
Chen, Yiming
Si, Lingyu
Zheng, Changwen
contents Scene understanding is one of the core tasks in computer vision, aiming to extract semantic information from images to identify objects, scene categories, and their interrelationships. Although advancements in Vision-Language Models (VLMs) have driven progress in this field, existing VLMs still face challenges in adaptation to unseen complex wide-area scenes. To address the challenges, this paper proposes a Hierarchical Coresets Selection (HCS) mechanism to advance the adaptation of VLMs in complex wide-area scene understanding. It progressively refines the selected regions based on the proposed theoretically guaranteed importance function, which considers utility, representativeness, robustness, and synergy. Without requiring additional fine-tuning, HCS enables VLMs to achieve rapid understandings of unseen scenes at any scale using minimal interpretable regions while mitigating insufficient feature density. HCS is a plug-and-play method that is compatible with any VLM. Experiments demonstrate that HCS achieves superior performance and universality in various tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2507_13061
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Advancing Complex Wide-Area Scene Understanding with Hierarchical Coresets Selection
Wang, Jingyao
Chen, Yiming
Si, Lingyu
Zheng, Changwen
Computer Vision and Pattern Recognition
Scene understanding is one of the core tasks in computer vision, aiming to extract semantic information from images to identify objects, scene categories, and their interrelationships. Although advancements in Vision-Language Models (VLMs) have driven progress in this field, existing VLMs still face challenges in adaptation to unseen complex wide-area scenes. To address the challenges, this paper proposes a Hierarchical Coresets Selection (HCS) mechanism to advance the adaptation of VLMs in complex wide-area scene understanding. It progressively refines the selected regions based on the proposed theoretically guaranteed importance function, which considers utility, representativeness, robustness, and synergy. Without requiring additional fine-tuning, HCS enables VLMs to achieve rapid understandings of unseen scenes at any scale using minimal interpretable regions while mitigating insufficient feature density. HCS is a plug-and-play method that is compatible with any VLM. Experiments demonstrate that HCS achieves superior performance and universality in various tasks.
title Advancing Complex Wide-Area Scene Understanding with Hierarchical Coresets Selection
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.13061