arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2310.07548cs.CV
收藏

属性定位与修正网络用于零样本学习

Attribute Localization and Revision Network for Zero-Shot Learning

Sichuan University(四川大学);Chengdu University of Information Technology(成都信息工程大学)

Junzhe Xu, Suling Duan, Chenwei Tang, Zhenan He, Jiancheng Lv

AI总结:

本文提出属性定位与修正网络,通过融合局部与全局特征并修正属性语义,解决零样本学习中属性对齐问题,在三个基准上验证了有效性。

摘要:

零样本学习使模型能够借助属性等辅助语义信息识别未见类别。现有工作提出从局部图像区域检测属性,并将提取的特征与类别级语义对齐。本文发现,局部特征与全局特征的选择并非零和博弈,全局特征同样有助于对属性的理解。此外,将属性特征与类别级语义对齐忽略了类内属性的潜在变化。为缓解这些不足,本文提出了属性定位与修正网络。首先,我们设计了属性定位模块(ALM),从图像区域捕获局部和全局特征,并引入一种名为尺度控制单元的新模块来融合全局与局部表示。其次,我们提出属性修正模块(ARM),通过修正每个属性的真实值生成图像级语义,补偿因忽略类内变化而导致的性能下降。最后,ALM的输出将与ARM生成的修正语义对齐,以完成训练过程。在三个广泛使用的基准上的综合实验结果表明,我们的模型在零样本预测任务中的有效性。

英文摘要:

Zero-shot learning enables the model to recognize unseen categories with the aid of auxiliary semantic information such as attributes. Current works proposed to detect attributes from local image regions and align extracted features with class-level semantics. In this paper, we find that the choice between local and global features is not a zero-sum game, global features can also contribute to the understanding of attributes. In addition, aligning attribute features with class-level semantics ignores potential intra-class attribute variation. To mitigate these disadvantages, we present Attribute Localization and Revision Network in this paper. First, we design Attribute Localization Module (ALM) to capture both local and global features from image regions, a novel module called Scale Control Unit is incorporated to fuse global and local representations. Second, we propose Attribute Revision Module (ARM), which generates image-level semantics by revising the ground-truth value of each attribute, compensating for performance degradation caused by ignoring intra-class variation. Finally, the output of ALM will be aligned with revised semantics produced by ARM to achieve the training process. Comprehensive experimental results on three widely used benchmarks demonstrate the effectiveness of our model in the zero-shot prediction task.

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。

↑