Zoo-Tuning: Adaptive Transfer from A Zoo of Models

Yang Shu, Zhi Kou, Zhangjie Cao, Jianmin Wang, Mingsheng Long
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:9626-9637, 2021.

Abstract

With the development of deep networks on various large-scale datasets, a large zoo of pretrained models are available. When transferring from a model zoo, applying classic single-model-based transfer learning methods to each source model suffers from high computational cost and cannot fully utilize the rich knowledge in the zoo. We propose \emph{Zoo-Tuning} to address these challenges, which learns to adaptively transfer the parameters of pretrained models to the target task. With the learnable channel alignment layer and adaptive aggregation layer, Zoo-Tuning \emph{adaptively aggregates channel aligned pretrained parameters to derive the target model}, which simultaneously promotes knowledge transfer and adapts source models to downstream tasks. The adaptive aggregation substantially reduces the computation cost at both training and inference. We further propose lite Zoo-Tuning with the temporal ensemble of batch average gating values to reduce the storage cost at the inference time. We evaluate our approach on a variety of tasks, including reinforcement learning, image classification, and facial landmark detection. Experiment results demonstrate that the proposed adaptive transfer learning approach can more effectively and efficiently transfer knowledge from a zoo of models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-shu21b, title = {Zoo-Tuning: Adaptive Transfer from A Zoo of Models}, author = {Shu, Yang and Kou, Zhi and Cao, Zhangjie and Wang, Jianmin and Long, Mingsheng}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {9626--9637}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/shu21b/shu21b.pdf}, url = {https://proceedings.mlr.press/v139/shu21b.html}, abstract = {With the development of deep networks on various large-scale datasets, a large zoo of pretrained models are available. When transferring from a model zoo, applying classic single-model-based transfer learning methods to each source model suffers from high computational cost and cannot fully utilize the rich knowledge in the zoo. We propose \emph{Zoo-Tuning} to address these challenges, which learns to adaptively transfer the parameters of pretrained models to the target task. With the learnable channel alignment layer and adaptive aggregation layer, Zoo-Tuning \emph{adaptively aggregates channel aligned pretrained parameters to derive the target model}, which simultaneously promotes knowledge transfer and adapts source models to downstream tasks. The adaptive aggregation substantially reduces the computation cost at both training and inference. We further propose lite Zoo-Tuning with the temporal ensemble of batch average gating values to reduce the storage cost at the inference time. We evaluate our approach on a variety of tasks, including reinforcement learning, image classification, and facial landmark detection. Experiment results demonstrate that the proposed adaptive transfer learning approach can more effectively and efficiently transfer knowledge from a zoo of models.} }
Endnote
%0 Conference Paper %T Zoo-Tuning: Adaptive Transfer from A Zoo of Models %A Yang Shu %A Zhi Kou %A Zhangjie Cao %A Jianmin Wang %A Mingsheng Long %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-shu21b %I PMLR %P 9626--9637 %U https://proceedings.mlr.press/v139/shu21b.html %V 139 %X With the development of deep networks on various large-scale datasets, a large zoo of pretrained models are available. When transferring from a model zoo, applying classic single-model-based transfer learning methods to each source model suffers from high computational cost and cannot fully utilize the rich knowledge in the zoo. We propose \emph{Zoo-Tuning} to address these challenges, which learns to adaptively transfer the parameters of pretrained models to the target task. With the learnable channel alignment layer and adaptive aggregation layer, Zoo-Tuning \emph{adaptively aggregates channel aligned pretrained parameters to derive the target model}, which simultaneously promotes knowledge transfer and adapts source models to downstream tasks. The adaptive aggregation substantially reduces the computation cost at both training and inference. We further propose lite Zoo-Tuning with the temporal ensemble of batch average gating values to reduce the storage cost at the inference time. We evaluate our approach on a variety of tasks, including reinforcement learning, image classification, and facial landmark detection. Experiment results demonstrate that the proposed adaptive transfer learning approach can more effectively and efficiently transfer knowledge from a zoo of models.
APA
Shu, Y., Kou, Z., Cao, Z., Wang, J. & Long, M.. (2021). Zoo-Tuning: Adaptive Transfer from A Zoo of Models. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:9626-9637 Available from https://proceedings.mlr.press/v139/shu21b.html.

Related Material