Report
MoTE: Reconciling Generalization with Specialization for Visual-Language to Video Knowledge Transfer
| العنوان: | MoTE: Reconciling Generalization with Specialization for Visual-Language to Video Knowledge Transfer |
|---|---|
| المؤلفون: | Zhu, Minghao, Wang, Zhengpu, Hu, Mengxian, Dang, Ronghao, Lin, Xiao, Zhou, Xun, Liu, Chengju, Chen, Qijun |
| سنة النشر: | 2024 |
| المجموعة: | Computer Science |
| مصطلحات موضوعية: | Computer Science - Computer Vision and Pattern Recognition |
| الوصف: | Transferring visual-language knowledge from large-scale foundation models for video recognition has proved to be effective. To bridge the domain gap, additional parametric modules are added to capture the temporal information. However, zero-shot generalization diminishes with the increase in the number of specialized parameters, making existing works a trade-off between zero-shot and close-set performance. In this paper, we present MoTE, a novel framework that enables generalization and specialization to be balanced in one unified model. Our approach tunes a mixture of temporal experts to learn multiple task views with various degrees of data fitting. To maximally preserve the knowledge of each expert, we propose \emph{Weight Merging Regularization}, which regularizes the merging process of experts in weight space. Additionally with temporal feature modulation to regularize the contribution of temporal feature during test. We achieve a sound balance between zero-shot and close-set video recognition tasks and obtain state-of-the-art or competitive results on various datasets, including Kinetics-400 \& 600, UCF, and HMDB. Code is available at \url{https://github.com/ZMHH-H/MoTE}. Comment: NeurIPS 2024 Camera Ready |
| نوع الوثيقة: | Working Paper |
| URL الوصول: | http://arxiv.org/abs/2410.10589 |
| رقم الانضمام: | edsarx.2410.10589 |
| قاعدة البيانات: | arXiv |
| الوصف غير متاح. |