Deep Residual Weight-Sharing Attention Network With Low-Rank Attention for Visual Question Answering

• Published in: IEEE transactions on multimedia Vol. 25; pp. 4282 - 4295
• Main Authors: Qin, Bosheng, Hu, Haoji, Zhuang, Yueting
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Ablation; Approximation algorithms; Feature extraction; Lightweight; low-rank attention; Mathematical models; Parameters; Qualitative analysis; Questions; residual learning; Routing; Statistical analysis; Task analysis; Transformers; Visual question answering; Visualization; Weight reduction; weight-sharing

The attention-based networks have become prevailing recently in visual question answering (VQA) due to their high performances. However, the extensive memory consumption of attention-based models poses excessive-high demand for the implementation equipment, raising concerns about their future applic...