MFL-AKD

数据集

1. Flickr30k

官网： https://shannon.cs.illinois.edu/DenotationGraph/

2. MSCOCO 5Fold 1K

官网： https://cocodataset.org/#home

3. Charades-STA

官网： https://prior.allenai.org/projects/charades

4. ActivityNet Captions

官网：https://cs.stanford.edu/people/ranjaykrishna/densevid/

按照： Qu, L., Liu, M., Wu, J., Gao, Z., & Nie, L. (2021, July). Dynamic modality interaction modeling for image-text retrieval. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 1104-1113).

Gao, J., Sun, C., Yang, Z., & Nevatia, R. (2017). Tall: Temporal activity localization via language query. In Proceedings of the IEEE international conference on computer vision (pp. 5267-5275).

设置（详见datasetting，或MFL-AKD中也有）

MFL-AKD

对比方法

单模态的方法按照FedAvg/MFL部署联邦学习

实用策略

调研时候用过的，可以作为联邦学习入门，因为很好操作