Interpreting Object-level Foundation Models via Visual Precision Search

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Ruoyu, Liang, Siyuan, Li, Jingzhi, Liu, Shiming, Li, Maosen, Huang, Zhen, Zhang, Hua, Cao, Xiaochun
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912308674953216
author Chen, Ruoyu
Liang, Siyuan
Li, Jingzhi
Liu, Shiming
Li, Maosen
Huang, Zhen
Zhang, Hua
Cao, Xiaochun
author_facet Chen, Ruoyu
Liang, Siyuan
Li, Jingzhi
Liu, Shiming
Li, Maosen
Huang, Zhen
Zhang, Hua
Cao, Xiaochun
contents Advances in multimodal pre-training have propelled object-level foundation models, such as Grounding DINO and Florence-2, in tasks like visual grounding and object detection. However, interpreting these models' decisions has grown increasingly challenging. Existing interpretable attribution methods for object-level task interpretation have notable limitations: (1) gradient-based methods lack precise localization due to visual-textual fusion in foundation models, and (2) perturbation-based methods produce noisy saliency maps, limiting fine-grained interpretability. To address these, we propose a Visual Precision Search method that generates accurate attribution maps with fewer regions. Our method bypasses internal model parameters to overcome attribution issues from multimodal fusion, dividing inputs into sparse sub-regions and using consistency and collaboration scores to accurately identify critical decision-making regions. We also conducted a theoretical analysis of the boundary guarantees and scope of applicability of our method. Experiments on RefCOCO, MS COCO, and LVIS show our approach enhances object-level task interpretability over SOTA for Grounding DINO and Florence-2 across various evaluation metrics, with faithfulness gains of 23.7%, 31.6%, and 20.1% on MS COCO, LVIS, and RefCOCO for Grounding DINO, and 102.9% and 66.9% on MS COCO and RefCOCO for Florence-2. Additionally, our method can interpret failures in visual grounding and object detection tasks, surpassing existing methods across multiple evaluation metrics. The code will be released at https://github.com/RuoyuChen10/VPS.
format Preprint
id arxiv_https___arxiv_org_abs_2411_16198
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Interpreting Object-level Foundation Models via Visual Precision Search
Chen, Ruoyu
Liang, Siyuan
Li, Jingzhi
Liu, Shiming
Li, Maosen
Huang, Zhen
Zhang, Hua
Cao, Xiaochun
Computer Vision and Pattern Recognition
Advances in multimodal pre-training have propelled object-level foundation models, such as Grounding DINO and Florence-2, in tasks like visual grounding and object detection. However, interpreting these models' decisions has grown increasingly challenging. Existing interpretable attribution methods for object-level task interpretation have notable limitations: (1) gradient-based methods lack precise localization due to visual-textual fusion in foundation models, and (2) perturbation-based methods produce noisy saliency maps, limiting fine-grained interpretability. To address these, we propose a Visual Precision Search method that generates accurate attribution maps with fewer regions. Our method bypasses internal model parameters to overcome attribution issues from multimodal fusion, dividing inputs into sparse sub-regions and using consistency and collaboration scores to accurately identify critical decision-making regions. We also conducted a theoretical analysis of the boundary guarantees and scope of applicability of our method. Experiments on RefCOCO, MS COCO, and LVIS show our approach enhances object-level task interpretability over SOTA for Grounding DINO and Florence-2 across various evaluation metrics, with faithfulness gains of 23.7%, 31.6%, and 20.1% on MS COCO, LVIS, and RefCOCO for Grounding DINO, and 102.9% and 66.9% on MS COCO and RefCOCO for Florence-2. Additionally, our method can interpret failures in visual grounding and object detection tasks, surpassing existing methods across multiple evaluation metrics. The code will be released at https://github.com/RuoyuChen10/VPS.
title Interpreting Object-level Foundation Models via Visual Precision Search
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2411.16198