AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

DR-Tune: Improving Fine-tuning of Pretrained Visual Models by Distribution Regularization with Semantic Calibration

Nan Zhou, Jiaxin Chen, Di Huang

ICCV 2023 · Architectures and efficiency › Pruning and efficient tuning · approach: Efficient ViT adaptation · 7 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
add distribution regularization on the task head using pretrained features (ablation)DR-Tune (CE-tuning base) · Caltech101top-1 accuracy (%): 91.93 -> 94.39Improved
add semantic calibration on top of distribution regularization (ablation)DR-Tune · Caltech101top-1 accuracy (%): 94.39 -> 95.73Improved
add distribution regularization alone (no semantic calibration) on fine-grained data (ablation)CE-tuning · Carstop-1 accuracy (%): 88.45 -> 89.03Mixed
add semantic calibration on top of DR (ablation)DR-Tune · Carstop-1 accuracy (%): 89.03 -> 90.60Improved
add global rotation to SC (ablation)DR-Tune SC module · Caltech101top-1 accuracy (%): 95.17 -> 95.73Improved
add class-level translation to SC (ablation)DR-Tune SC module · Caltech101top-1 accuracy (%): 95.59 -> 95.73Improved
add confidence guided average for class centers (ablation)DR-Tune SC module · Petstop-1 accuracy (%): 89.69 -> 90.24Improved
class-level rotation instead of global rotation (ablation)DR-Tune SC module · ImageNet20top-1 accuracy (%): 96.03 (GR) vs 95.82 (CLR)No clear effect
apply DR-Tune instead of CE-tuning across pretraining strategiesResNet-50 SimSiam · Caltech101top-1 accuracy (%): 82.28 -> 90.33Improved
reduce fine-tuning data to 10% of ImageNet20DR-Tune vs CE-tuning · ImageNet20top-1 accuracy drop: DR-Tune drop 3.3% vs CE-tuning drop 29.9%Improved
use DR-Tune instead of knowledge distillation regularizersResNet-50 student · DTDtop-1 accuracy (%): 76.86 (best KD) -> 77.97Improved
use DR-Tune with ResNet-152 backboneDR-Tune · DTDtop-1 accuracy (%): 70.16 -> 71.44Improved

Later papers that compared with this method

1 later accepted paper reports DR-Tune as a baseline in their results tables; 0 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