2020/11/01 by Simon Ging, Ging, Simon, Mohammadreza Zolfaghari +5 · 3 citations
Computer Science · #Human Pose and Action Recognition #Multimodal Machine Learning Applications #Video Analysis and Summarization #cs.AI #cs.CL #cs.CV #cs.LG
paper · pdf · doi:10.48550/arxiv.2011.00597
27 pages, 5 figures, 19 tables. To be published in the 34th conference on Neural Information Processing Systems (NeurIPS 2020). The first two authors contributed equally to this work
arxiv created 2020/11/01 · arxiv updated 2020/11/03
Many real-world video-text tasks involve different levels of granularity, such as frames and words, clip and sentences or videos and paragraphs, each with distinct semantics. In this paper, we propose a Cooperative hierarchical Transformer (COOT) to leverage this hierarchy information and model the interactions between different levels of granularity and different modalities. The method consists of three major components: an attention-aware feature aggregation layer, which leverages the local temporal context (intra-level, e.g., within a clip), a contextual transformer to learn the interactions between low-level and high-level semantics (inter-level, e.g. clip-video, sentence-paragraph), and a cross-modal cycle-consistency loss to connect video and text. The resulting method compares favorably to the state of the art on several benchmarks while having few parameters. All code is available open-source at https://github.com/gingsi/coot-videotext