AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Zoo-Tuning: Adaptive Transfer from A Zoo of Models

Yang Shu, Zhi Kou, Zhangjie Cao, Jianmin Wang, Mingsheng Long

ICML 2021 · Architectures and efficiency › Pruning and efficient tuning · approach: Low-rank compression and merging · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
adaptive aggregation (Zoo-Tuning) of source models vs single-model fine-tuningResNet-50 model zoo · 7 classification tasksAvg. Acc.: 85.16 vs 83.41 best single (ImageNet sup.)Improved
temporal-ensemble unified gating (lite Zoo-Tuning) instead of per-sample gatingZoo-Tuning · 7 classification tasksAvg. Acc.: 84.51 vs 85.16Hurt
removing channel alignment layer T (ablation)Zoo-Tuning · COCO-70accuracy: 83.92 vs 84.91Hurt
removing channel alignment layer T (ablation)Zoo-Tuning · WFLWNME: 4.64 vs 4.58Hurt
learned adaptive gating vs average aggregation (ablation)Zoo-Tuning · COCO-70accuracy: 84.91 vs 80.53Improved
learned adaptive gating vs average aggregation (ablation)Zoo-Tuning · WFLWNME: 4.58 vs 4.75Improved
5 ImageNet supervised models aggregated instead of 5 diverse source models (same parameter count) (ablation)Zoo-Tuning · Aircraft / COCO-70accuracy: 84.92 vs 86.54 (Aircraft); 82.13 vs 84.91 (COCO)Hurt
increasing number of source models from 2 to 5 (ablation)Zoo-Tuning · CIFAR-100accuracy: 78.62 (2 models) to 83.77 (5 models)Improved
fine-tuning each source model separately (no aggregation) on RL target taskssingle-model transfer in Atari RL · Atari transfer (Alien, Gopher, JamesBond)scoreNo clear effect
Zoo-Tuning vs Knowledge Flow on facial landmarksmulti-model transfer · 300W/WFLW/COFWNME: 4.58 vs 5.28 on WFLWImproved
Zoo-Tuning vs ensemble on facial landmarksmulti-model transfer · 300W/COFWNME: 3.41 vs 3.33 (300W); 3.51 vs 3.46 (COFW)Hurt

Later papers that compared with this method

1 later accepted paper reports Zoo Tuning as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy