arXiv 2510.105872025-10-12cs.CV

视觉定位的一个简单且更好的基线方法

A Simple and Better Baseline for Visual Grounding

School of Data Science（数据科学学院）；Engineering East China Normal University Shanghai, China（工程东华师范大学上海中国）；OpenScience Lab Shanghai AI Laboratory Shanghai, China（OpenScience Lab 上海AI实验室上海中国）；School of Life Science（生命科学学院）；Technology Xi'an Jiaotong University Xi'an, China（技术西安交通大学西安中国）；Medical Artificial Intelligence Laboratory Westlake University Hangzhou, China（医学人工智能实验室西湖大学杭州中国）

Jingchao Wang, Wenlong Zhang, Dingjiang Huang, Hong Wang, Yefeng Zheng

备注：: ICME2025

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出了一种基于特征选择的简单有效视觉定位基线FSVG，通过整体网络架构并行利用语言引导模态交互，并引入相似度特征选择机制降低计算成本，在准确率与效率上超越现有方法。

摘要：

视觉定位旨在预测由文本描述指定的目标对象的位置。对于这种涉及语言和视觉模态的任务，最新的研究路线专注于仅选择与语言相关的视觉区域进行对象定位，以减少计算开销。尽管取得了令人瞩目的性能，但该方法在不同图像尺度上迭代执行，且每次迭代时，语言特征和视觉特征都需要存储在缓存中，从而产生额外开销。为了便于实现，本文提出了一种基于特征选择的简单而有效的视觉定位基线方法，称为FSVG。具体而言，我们直接将语言和视觉模态封装到一个整体网络架构中，无需复杂的迭代过程，并并行利用语言作为引导，促进语言模态和视觉模态之间的交互，以提取有效的视觉特征。此外，为了降低计算成本，在视觉特征学习期间，我们引入了一种基于相似度的特征选择机制，仅利用与语言相关的视觉特征以实现更快的预测。在多个基准数据集上进行的广泛实验全面证实，所提出的FSVG在准确性和效率之间取得了超越当前最先进方法的更好平衡。代码可在https://github.com/jcwang0602/FSVG获取。

英文摘要：

Visual grounding aims to predict the locations of target objects specified by textual descriptions. For this task with linguistic and visual modalities, there is a latest research line that focuses on only selecting the linguistic-relevant visual regions for object localization to reduce the computational overhead. Albeit achieving impressive performance, it is iteratively performed on different image scales, and at every iteration, linguistic features and visual features need to be stored in a cache, incurring extra overhead. To facilitate the implementation, in this paper, we propose a feature selection-based simple yet effective baseline for visual grounding, called FSVG. Specifically, we directly encapsulate the linguistic and visual modalities into an overall network architecture without complicated iterative procedures, and utilize the language in parallel as guidance to facilitate the interaction between linguistic modal and visual modal for extracting effective visual features. Furthermore, to reduce the computational cost, during the visual feature learning, we introduce a similarity-based feature selection mechanism to only exploit language-related visual features for faster prediction. Extensive experiments conducted on several benchmark datasets comprehensively substantiate that the proposed FSVG achieves a better balance between accuracy and efficiency beyond the current state-of-the-art methods. Code is available at https://github.com/jcwang0602/FSVG.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。