Exploring Sparse Spatial Relation in Graph Inference for Text-Based VQA

• Published in: IEEE transactions on image processing Vol. 32; pp. 5060 - 5074
• Main Authors: Zhou, Sheng, Guo, Dan, Li, Jia, Yang, Xun, Wang, Meng
• Format: Journal Article
• Language: English
• Published: New York IEEE 2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Cognition; Computational modeling; graph inference; Inference; Learning; Optical character recognition; Predictions; Pruning; Question answering (information retrieval); Reasoning; relation learning; spatial relation; Task analysis; text-based visual question answering; Transformers; Visual observation; Visual question answering; Visualization

Text-based visual question answering (TextVQA) faces the significant challenge of avoiding redundant relational inference. To be specific, a large number of detected objects and optical character recognition (OCR) tokens result in rich visual relationships. Existing works take all visual relationshi...