CogTree: Cognition Tree Loss for Unbiased Scene Graph Generation

Author:

Yu Jing1,Chai Yuan2,Wang Yujing3,Hu Yue4,Wu Qi5

Affiliation:

1. Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China

2. Intelligent Computing & Machine Learning Lab, School of ASEE, Beihang University, Beijing, China

3. Key Laboratory of Machine Perception, MOE, School of EECS, Peking University, Beijing, China

4. Institute of Information Engineering,Chinese Academy of Sciences, Beijing, China

5. University of Adelaide, Australia

Abstract

Scene graphs are semantic abstraction of images that encourage visual understanding and reasoning. However, the performance of Scene Graph Generation (SGG) is unsatisfactory when faced with biased data in real-world scenarios. Conventional debiasing research mainly studies from the view of balancing data distribution or learning unbiased models and representations, ignoring the correlations among the biased classes. In this work, we analyze this problem from a novel cognition perspective: automatically building a hierarchical cognitive structure from the biased predictions and navigating that hierarchy to locate the relationships, making the tail relationships receive more attention in a coarse-to-fine mode. To this end, we propose a novel debiasing Cognition Tree (CogTree) loss for unbiased SGG. We first build a cognitive structure CogTree to organize the relationships based on the prediction of a biased SGG model. The CogTree distinguishes remarkably different relationships at first and then focuses on a small portion of easily confused ones. Then, we propose a debiasing loss specially for this cognitive structure, which supports coarse-to-fine distinction for the correct relationships. The loss is model-agnostic and consistently boosting the performance of several state-of-the-art models. The code is available at: https://github.com/CYVincent/Scene-Graph-Transformer-CogTree.

Publisher

International Joint Conferences on Artificial Intelligence Organization

Cited by 54 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Relation-Specific Feature Augmentation for unbiased scene graph generation;Pattern Recognition;2025-01

2. NICEST: Noisy Label Correction and Training for Robust Scene Graph Generation;IEEE Transactions on Pattern Analysis and Machine Intelligence;2024-10

3. Enhanced Data Transfer Cooperating with Artificial Triplets for Scene Graph Generation;IEICE Transactions on Information and Systems;2024-09-01

4. Review on scene graph generation methods;Multiagent and Grid Systems;2024-08-12

5. From Easy to Hard: Learning Curricular Shape-Aware Features for Robust Panoptic Scene Graph Generation;International Journal of Computer Vision;2024-08-05

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3