Improving Interpretable Embeddings for Ad-hoc Video Search with Generative Captions and Multi-word Concept Bank
This work addresses dataset and concept bank bottlenecks for researchers and practitioners in video search, though it is incremental as it enhances an existing state-of-the-art method.
The paper tackled the limitations in ad-hoc video search caused by small datasets and low-quality concept banks by constructing a new 7 million generated text-video dataset and a multi-word concept bank, resulting in doubled R@1 performance on MSRVTT and average 20% improvement in xinfAP on TRECVid AVS queries over eight years.
Aligning a user query and video clips in cross-modal latent space and that with semantic concepts are two mainstream approaches for ad-hoc video search (AVS). However, the effectiveness of existing approaches is bottlenecked by the small sizes of available video-text datasets and the low quality of concept banks, which results in the failures of unseen queries and the out-of-vocabulary problem. This paper addresses these two problems by constructing a new dataset and developing a multi-word concept bank. Specifically, capitalizing on a generative model, we construct a new dataset consisting of 7 million generated text and video pairs for pre-training. To tackle the out-of-vocabulary problem, we develop a multi-word concept bank based on syntax analysis to enhance the capability of a state-of-the-art interpretable AVS method in modeling relationships between query words. We also study the impact of current advanced features on the method. Experimental results show that the integration of the above-proposed elements doubles the R@1 performance of the AVS method on the MSRVTT dataset and improves the xinfAP on the TRECVid AVS query sets for 2016-2023 (eight years) by a margin from 2% to 77%, with an average about 20%.