Motivated by the success of meta-learning, we use the validation set to update the meta self-paced network during the training of the matching network.
Oct 24, 2021 · In this paper, we propose a Meta Self-Paced Network. (Meta-SPN) that automatically learns a weighting scheme from data for cross-modal matching.
Meta Self-Paced Learning for Cross-Modal Matching. https://doi.org/10.1145/3474085.3475451. Journal: Proceedings of the 29th ACM International Conference on ...
Meta Self-Paced Learning for Cross-Modal Matching · Jiwei Wei, Xing Xu, Zheng Wang, Guoqing Wang. 2021 (modified: 18 Nov 2022); ACM Multimedia 2021; Readers ...
... Cross-modal Alignment Network for Text-Video Retrieval. ACM Multimedia, 2021. [paper]; [Wei et al. ACMMM21] Meta Self-Paced Learning for Cross-Modal Matching.
(CVPR2020_MPL) Universal Weighting Metric Learning for Cross-Modal Matching. ... (ACMMM2021_Meta-SPN) Meta Self-Paced Learning for Cross-Modal Matching.
Meta self-paced learning for cross-modal matching. J Wei, X Xu, Z Wang, G Wang. Proceedings of the 29th ACM international conference on multimedia, 3835-3843 ...
... Meta Self-Paced Learning for Cross-Modal Matching. ACM International Conference on Multimedia 2021 (CCF A, Corresponding author). Jingran Zhang, Xing Xu ...
Multi-modal learning is a technique that builds classification/segmentation models that integrate and process information from multiple data modalities (e.g., ...
Mar 5, 2024 · We pre-trained a generative model, named SpeechFlow, on 60k hours of untranscribed speech with Flow Matching and masked conditions.