Fine-Grained Scene Graph Generation with Data Transfer

• Published in: Computer Vision - ECCV 2022 Vol. 13687; pp. 409 - 424
• Main Authors: Zhang, Ao, Yao, Yuan, Chen, Qianyu, Ji, Wei, Liu, Zhiyuan, Sun, Maosong, Chua, Tat-Seng
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer 2022; Springer Nature Switzerland
• Series: Lecture Notes in Computer Science
• Subjects: Large-scale; Plug-and-play; Scene graph generation
• ISBN: 9783031198113; 3031198115
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-031-19812-0_24

Scene graph generation (SGG) is designed to extract (subject, predicate, object) triplets in images. Recent works have made a steady progress on SGG, and provide useful tools for high-level vision and language understanding. However, due to the data distribution problems including long-tail distribution and semantic ambiguity, the predictions of current SGG models tend to collapse to several frequent but uninformative predicates (e.g., on, at), which limits practical application of these models in downstream tasks. To deal with the problems above, we propose a novel Internal and External Data Transfer (IETrans) method, which can be applied in a plug-and-play fashion and expanded to large SGG with 1,807 predicate classes. Our IETrans tries to relieve the data distribution problem by automatically creating an enhanced dataset that provides more sufficient and coherent annotations for all predicates. By applying our proposed method, a Neural Motif model doubles the macro performance for informative SGG. The code and data are publicly available at https://github.com/waxnkw/IETrans-SGG.pytorch.