Data

Tools

Indexes

Applications

Experiments

Open Source Science

regression-tta

The official implementation of "Test-time Adaptation for Regression by Subspace Alignment" (ICLR 2025).

https://github.com/kzkadc/regression-tta

Science Score: 36.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ✓
    Academic publication links
    Links to: arxiv.org
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (10.0%) to scientific vocabulary

Keywords

deep-learning pytorch regression test-time-adaptation
Last synced: 6 months ago · JSON representation

Repository

The official implementation of "Test-time Adaptation for Regression by Subspace Alignment" (ICLR 2025).

Basic Info
  • Host: GitHub
  • Owner: kzkadc
  • License: other
  • Language: Python
  • Default Branch: main
  • Homepage:
  • Size: 45.8 MB
Statistics
  • Stars: 4
  • Watchers: 1
  • Forks: 2
  • Open Issues: 0
  • Releases: 0
Topics
deep-learning pytorch regression test-time-adaptation
Created about 1 year ago · Last pushed 9 months ago
Metadata Files
Readme License Citation

README.md

Test-time Adaptation for Regression by Subspace Alignment (ICLR 2025)

The official implementation for ICLR 2025 paper "Test-time Adaptation for Regression by Subspace Alignment."
[OpenReview] [arXiv] [Poster]

Abstract This paper investigates test-time adaptation (TTA) for regression, where a regression model pre-trained in a source domain is adapted to an unknown target distribution with unlabeled target data. Although regression is one of the fundamental tasks in machine learning, most of the existing TTA methods have classification-specific designs, which assume that models output class-categorical predictions, whereas regression models typically output only single scalar values. To enable TTA for regression, we adopt a feature alignment approach, which aligns the feature distributions between the source and target domains to mitigate the domain gap. However, we found that naive feature alignment employed in existing TTA methods for classification is ineffective or even worse for regression because the features are distributed in a small subspace and many of the raw feature dimensions have little significance to the output. For an effective feature alignment in TTA for regression, we propose Significant-subspace Alignment (SSA). SSA consists of two components: subspace detection and dimension weighting. Subspace detection finds the feature subspace that is representative and significant to the output. Then, the feature alignment is performed in the subspace during TTA. Meanwhile, dimension weighting raises the importance of the dimensions of the feature subspace that have greater significance to the output. We experimentally show that SSA outperforms various baselines on real-world datasets.

overview

0. Environment

  • Prepare the datasets (SVHN, MNIST, UTKFace, Biwi Kinect, California Housing) and write their paths in dataset/dataset_config.py.
  • Install dependencies or build the docker image according to docker/Dockerfile.

bash $ docker build -t tta_regression docker --no-cache

1. Training the source model

```bash $ python3 trainsource.py -c configs/trainsource/svhn.yaml -o result/source/svhn

running with the docker image

$ docker run -it --rm -v $(pwd):$(pwd) -w $(pwd) --gpus device=0 ttaregression python3 trainsource.py -c configs/train_source/svhn.yaml -o result/source/svhn ```

2. Computing the feature statistics

```bash $ python3 featurestats.py -c configs/featurestats/svhn.yaml -o result/source/svhn

running with the docker image

$ docker run -it --rm -v $(pwd):$(pwd) -w $(pwd) --gpus device=0 ttaregression python3 featurestats.py -c configs/feature_stats/svhn.yaml -o result/source/svhn ```

The pre-trained model and feature statistics for SVHN are available in result/.

3. TTA

```bash $ python3 adaptation.py -c configs/tta/svhn.yaml -o result/tta/svhn

running with the docker image

$ docker run -it --rm -v $(pwd):$(pwd) -w $(pwd) --gpus device=0 tta_regression python3 adaptation.py -c configs/tta/svhn.yaml -o result/tta/svhn ```

Citation

If our work assists your research, please cite our paper:

@inproceedings{adachi2025testtime, title={Test-time Adaptation for Regression by Subspace Alignment}, author={Kazuki Adachi and Shin'ya Yamaguchi and Atsutoshi Kumagai and Tomoki Hamagami}, booktitle={The Thirteenth International Conference on Learning Representations}, year={2025}, url={https://openreview.net/forum?id=SXtl7NRyE5} }

Owner

  • Name: Kazuki Adachi
  • Login: kzkadc
  • Kind: user
  • Location: Japan
  • Company: NTT

GitHub Events

Total
  • Watch event: 9
  • Issue comment event: 1
  • Push event: 11
  • Fork event: 3
  • Create event: 1
Last Year
  • Watch event: 9
  • Issue comment event: 1
  • Push event: 11
  • Fork event: 3
  • Create event: 1

Issues and Pull Requests

Last synced: 12 months ago

All Time
  • Total issues: 0
  • Total pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Total issue authors: 0
  • Total pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 0
  • Pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Issue authors: 0
  • Pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
View more stats
Top Authors
Issue Authors
Pull Request Authors
Top Labels
Issue Labels
Pull Request Labels

Dependencies

docker/Dockerfile docker
  • nvidia/cuda 12.6.3-cudnn-devel-ubuntu24.04 build