Saved in:
Bibliographic Details
Main Authors: Gu, Zihan, Chen, Ruoyu, Zhang, Junchi, Hu, Yue, Zhang, Hua, Cao, Xiaochun
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.10914
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911264631947264
author Gu, Zihan
Chen, Ruoyu
Zhang, Junchi
Hu, Yue
Zhang, Hua
Cao, Xiaochun
author_facet Gu, Zihan
Chen, Ruoyu
Zhang, Junchi
Hu, Yue
Zhang, Hua
Cao, Xiaochun
contents Attribution is essential for interpreting object-level foundation models. Recent methods based on submodular subset selection have achieved high faithfulness, but their efficiency limitations hinder practical deployment in real-world scenarios. To address this, we propose PhaseWin, a novel phase-window search algorithm that enables faithful region attribution with near-linear complexity. PhaseWin replaces traditional quadratic-cost greedy selection with a phased coarse-to-fine search, combining adaptive pruning, windowed fine-grained selection, and dynamic supervision mechanisms to closely approximate greedy behavior while dramatically reducing model evaluations. Theoretically, PhaseWin retains near-greedy approximation guarantees under mild monotone submodular assumptions. Empirically, PhaseWin achieves over 95% of greedy attribution faithfulness using only 20% of the computational budget, and consistently outperforms other attribution baselines across object detection and visual grounding tasks with Grounding DINO and Florence-2. PhaseWin establishes a new state of the art in scalable, high-faithfulness attribution for object-level multimodal models.
format Preprint
id arxiv_https___arxiv_org_abs_2511_10914
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PhaseWin Search Framework Enable Efficient Object-Level Interpretation
Gu, Zihan
Chen, Ruoyu
Zhang, Junchi
Hu, Yue
Zhang, Hua
Cao, Xiaochun
Computer Vision and Pattern Recognition
Attribution is essential for interpreting object-level foundation models. Recent methods based on submodular subset selection have achieved high faithfulness, but their efficiency limitations hinder practical deployment in real-world scenarios. To address this, we propose PhaseWin, a novel phase-window search algorithm that enables faithful region attribution with near-linear complexity. PhaseWin replaces traditional quadratic-cost greedy selection with a phased coarse-to-fine search, combining adaptive pruning, windowed fine-grained selection, and dynamic supervision mechanisms to closely approximate greedy behavior while dramatically reducing model evaluations. Theoretically, PhaseWin retains near-greedy approximation guarantees under mild monotone submodular assumptions. Empirically, PhaseWin achieves over 95% of greedy attribution faithfulness using only 20% of the computational budget, and consistently outperforms other attribution baselines across object detection and visual grounding tasks with Grounding DINO and Florence-2. PhaseWin establishes a new state of the art in scalable, high-faithfulness attribution for object-level multimodal models.
title PhaseWin Search Framework Enable Efficient Object-Level Interpretation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.10914