VALSE 论文速览 第166期：面向细粒度场景图生成的环境不变课程关系学习 ...

论文题目：

Environment-Invariant Curriculum Relation Learning for Fine-Grained Scene Graph Generation

作者列表：

闵聿宽 (西安电子科技大学)、武阿明 (西安电子科技大学)、邓成 (西安电子科技大学)

B站观看网址：

https://www.bilibili.com/video/BV1YK421y7XE/

论文摘要：

The scene graph generation (SGG) task is designed to identify the predicates based on the subject-object pairs. However , existing datasets generally include two imbalance cases: one is the class imbalance from the predicted predicates and another is the context imbalance from the given subject-object pairs, which presents significant challenges for SGG. Most existing methods focus on the imbalance of the predicted predicate while ignoring the imbalance of the subject-object pairs, which could not achieve satisfactory results. To address the two imbalance cases, we propose a novel Environment Invariant Curriculum Relation learning (EICR) method, which can be applied in a plug-and-play fashion to existing SGG methods. Concretely, to remove the imbalance of the subject-object pairs, we first construct different distribution environments for the subject-object pairs and learn a model invariant to the environment changes. Then, we construct a class-balanced curriculum learning strategy to balance the different environments to remove the predicate imbalance. Comprehensive experiments conducted on VG and GQA datasets demonstrate that our EICR framework can be taken as a general strategy for various SGG models, and achieve significant improvements.

参考文献：

[1] Yukuan Min, Aming Wu, and Cheng Deng “Environment-Invariant Curriculum Relation Learning for Fine-Grained Scene Graph Generation”, in Proceeding of IEEE International Conference on Computer Vision (ICCV2023), Paris, France, Oct 2023.

论文链接：

[https://openaccess.thecvf.com/content/ICCV2023/papers/Min_Environment-Invariant_Curriculum_Relation_Learning_for_Fine-Grained_Scene_Graph_Generation_ICCV_2023_paper.pdf]

代码链接：

[https://github.com/myukzzz/EICR]

视频讲者简介：

闵聿宽，西安电子科技大学电子工程学院博士研究生，师从邓成教授，主要研究方向是计算机视觉、场景图生成和深度学习。

特别鸣谢本次论文速览主要组织者：

月度轮值AC：胡迪 (中国人民大学)