Learning Dual Low-Rank Representation for Multi-Label Micro-Video Classification
Autor: | Yuting Su, Li Desheng, Peiguang Jing, Liqiang Nie, Wei Lu |
---|---|
Rok vydání: | 2023 |
Předmět: |
Modality (human–computer interaction)
business.industry Computer science Rank (computer programming) Construct (python library) DUAL (cognitive architecture) Semantics Machine learning computer.software_genre Computer Science Applications Consistency (database systems) Categorization Signal Processing Media Technology Artificial intelligence Electrical and Electronic Engineering business Representation (mathematics) computer |
Zdroj: | IEEE Transactions on Multimedia. 25:77-89 |
ISSN: | 1941-0077 1520-9210 |
Popis: | Currently, with the rapid development of mobile Internet, micro-video has become a prevailing format of user-generated contents (UGCs) on various social media platforms. Several studies have been conducted towards to understanding high-level micro-video semantics, such as venue categorization, memorability analysis, and popularity prediction. However, these approaches supported tasks with only a single output, which exhibited limitations when attempting to use them to resolve tasks with multiple outputs, especially the multi-label micro-video classification. To tackle this problem, in this paper, we propose a dual multi-modal low-rank decomposition (DMLRD) method for multi-label micro-video classification tasks. To learn more comprehensive micro-video representations, we first learn the low-rank-regularized modality-specific and modality-shared components by considering the consistency and the complementarity among modalities simultaneously. Meanwhile, the less descriptive power of each modality aroused by inherent properties can be solved. To obtain unseen label representations, we next construct a sparsity-regularized multi-matrix normal estimation term to jointly encode the latent relationship structures among labels and dimensions. Experiments on a publicly available large-scale micro-video dataset demonstrate the effectiveness of our proposed model over the state-of-art methods. |
Databáze: | OpenAIRE |
Externí odkaz: |