Learning Two-Branch Neural Networks for Image-Text Matching Tasks

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 41; no. 2; pp. 394 - 407
• Main Authors: Wang, Liwei, Li, Yin, Huang, Jing, Lazebnik, Svetlana
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.02.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Bidirectional control; Computer vision; cross-modal retrieval; Deep learning; Embedding; Feature extraction; Fuses; Grounding; image-sentence retrieval; Matching; Natural languages; Neighborhoods; Neural networks; phrase localization; Sampling; Sentences; Similarity; Task analysis; Training; visual grounding; Visualization

Image-language matching tasks have recently attracted a lot of attention in the computer vision field. These tasks include image-sentence matching, i.e., given an image query, retrieving relevant sentences and vice versa, and region-phrase matching or visual grounding, i.e., matching a phrase to rel...