Video-Text Retrieval by Supervised Sparse Multi-Grained Learning

19 Feb 2023  ·  Yimu Wang, Peng Shi ·

While recent progress in video-text retrieval has been advanced by the exploration of better representation learning, in this paper, we present a novel multi-grained sparse learning framework, S3MA, to learn an aligned sparse space shared between the video and the text for video-text retrieval. The shared sparse space is initialized with a finite number of sparse concepts, each of which refers to a number of words. With the text data at hand, we learn and update the shared sparse space in a supervised manner using the proposed similarity and alignment losses. Moreover, to enable multi-grained alignment, we incorporate frame representations for better modeling the video modality and calculating fine-grained and coarse-grained similarities. Benefiting from the learned shared sparse space and multi-grained similarities, extensive experiments on several video-text retrieval benchmarks demonstrate the superiority of S3MA over existing methods. Our code is available at https://github.com/yimuwangcs/Better_Cross_Modal_Retrieval.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Video Retrieval MSR-VTT-1kA SuMA (ViT-B/16) text-to-video R@1 49.8 # 16
text-to-video R@5 75.1 # 20
text-to-video R@10 83.9 # 19
video-to-text R@1 47.3 # 15
video-to-text R@5 76 # 8
video-to-text R@10 84.3 # 11

Methods


No methods listed for this paper. Add relevant methods here