Object-aware Video-language Pre-training for Retrieval
This work addresses the challenge of fine-grained semantic alignment in video-language retrieval for AI applications, representing an incremental improvement over existing transformer models.
The paper tackles the problem of fine-grained semantic alignment in video-language pre-training by proposing Object-aware Transformers, which incorporate object representations using bounding boxes and object tags. The result shows clear improvements in performance across all tasks and datasets considered, with specific gains on three standard sub-tasks of video-text matching across four benchmarks.
Recently, by introducing large-scale dataset and strong transformer network, video-language pre-training has shown great success especially for retrieval. Yet, existing video-language transformer models do not explicitly fine-grained semantic align. In this work, we present Object-aware Transformers, an object-centric approach that extends video-language transformer to incorporate object representations. The key idea is to leverage the bounding boxes and object tags to guide the training process. We evaluate our model on three standard sub-tasks of video-text matching on four widely used benchmarks. We also provide deep analysis and detailed ablation about the proposed method. We show clear improvement in performance across all tasks and datasets considered, demonstrating the value of a model that incorporates object representations into a video-language architecture. The code will be released at \url{https://github.com/FingerRec/OA-Transformer}.