Timezone: »

ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning
Junting Pan · Ziyi Lin · Xiatian Zhu · Jing Shao · Hongsheng Li

Wed Dec 07 05:00 PM -- 07:00 PM (PST) @

Capitalizing on large pre-trained models for various downstream tasks of interest have recently emerged with promising performance. Due to the ever-growing model size, the standard full fine-tuning based task adaptation strategy becomes prohibitively costly in terms of model training and storage. This has led to a new research direction in parameter-efficient transfer learning. However, existing attempts typically focus on downstream tasks from the same modality (e.g., image understanding) of the pre-trained model. This creates a limit because in some specific modalities, (e.g., video understanding) such a strong pre-trained model with sufficient knowledge is less or not available. In this work, we investigate such a novel cross-modality transfer learning setting, namely parameter-efficient image-to-video transfer learning. To solve this problem, we propose a new Spatio-Temporal Adapter (ST-Adapter) for parameter-efficient fine-tuning per video task. With a built-in spatio-temporal reasoning capability in a compact design, ST-Adapter enables a pre-trained image model without temporal knowledge to reason about dynamic video content at a small ~8% per-task parameter cost, requiring approximately 20 times fewer updated parameters compared to previous work. Extensive experiments on video action recognition tasks show that our ST-Adapter can match or even outperform the strong full fine-tuning strategy and state-of-the-art video models, whilst enjoying the advantage of parameter efficiency.

Author Information

Junting Pan (The Chinese University of Hong Kong)
Ziyi Lin (The Chinese University of Hong Kong)
Xiatian Zhu (University of Surrey)
Jing Shao (Sensetime)
Hongsheng Li (The Chinese University of Hong Kong)

More from the Same Authors