Contrastive video question answering via video graph transformer

• Main Authors: XIAO, Junbin Xiao, ZHOU, Pan, YAO, Angela, LI, Yicong, HONG, Richang, YAN, Shuicheng, CHUA, Tat-Seng
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2023
• Subjects: VideoQA; Cross-Modal Visual Reasoning; Video-Language; Dynamic Visual Graphs; Contrastive Learning; Transformer; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/9053; https://ink.library.smu.edu.sg/context/sis_research/article/10056/viewcontent/2023_TPAMI_ContrastiveVideo.pdf
• Institution: Singapore Management University

We propose to perform video question answering (VideoQA) in a Contrastive manner via a Video Graph Transformer model (CoVGT). CoVGT’s uniqueness and superiority are three-fold: 1) It proposes a dynamic graph transformer module which encodes video by explicitly capturing the visual objects, their relations and dynamics, for complex spatio-temporal reasoning. 2) It designs separate video and text transformers for contrastive learning between the video and text to perform QA, instead of multi-modal transformer for answer classification. Fine-grained video-text communication is done by additional cross-modal interaction modules. 3) It is optimized by the joint fully- and self-supervised contrastive objectives between the correct and incorrect answers, as well as the relevant and irrelevant questions respectively. With superior video encoding and QA solution, we show that CoVGT can achieve much better performances than previous arts on video reasoning tasks. Its performances even surpass those models that are pretrained with millions of external data. We further show that CoVGT can also benefit from cross-modal pretraining, yet with orders of magnitude smaller data. The results demonstrate the effectiveness and superiority of CoVGT, and additionally reveal its potential for more data-efficient pretraining. We hope our success can advance VideoQA beyond coarse recognition/description towards fine-grained relation reasoning of video contents. Our code is available at https://github.com/doc-doc/CoVGT.