Research map / Label-efficient and robust learning / Domain adaptation and generalization
PUnDA: Probabilistic Unsupervised Domain Adaptation for Knowledge Transfer Across Visual Categories
Behnam Gholami, Ognjen (Oggi) Rudovic, Vladimir Pavlovic
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace ILS with PUnDA (probabilistic joint latent space and classifier) | unsupervised DA · Office+Caltech10 average, VGG-FC6 features | accuracy: 81.3 -> 83.4 | Improved |
| replace ILS with PUnDA | unsupervised DA · Office+Caltech10 average, VGG-FC7 features | accuracy: 79.2 -> 81.7 | Improved |
| replace ILS with PUnDA | unsupervised DA · Office+Caltech10 average, SURF features | accuracy: 46.7 -> 48.8 | Improved |
| replace ILS with PUnDA | unsupervised DA · Multi-PIE average (source C27) | accuracy: 64.2 -> 65.7 | Improved |
| replace ILS with PUnDA on small-angle transfer | unsupervised DA · Multi-PIE C09 (source C27) | accuracy: 96.6 -> 94.3 | Hurt |
| add domain mismatch regularization term (lambda greater than 0) (ablation) | PUnDA · Office+Caltech10, VGG-FC6 | accuracy | Improved |
| vary lambda' over 0 to 10 (ablation) | PUnDA · Office+Caltech10, VGG-FC6 | accuracy | No clear effect |
| use unlabeled target data in classifier learning (ablation) | PUnDA · Office+Caltech10, VGG-FC6 | accuracy | Improved |
| use VGG-FC7 instead of VGG-FC6 features | PUnDA · Office+Caltech10 average | accuracy: 83.4 -> 81.7 | Hurt |
Later papers that compared with this method
1 later accepted paper reports PU-NDA as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Leveraging Contaminated Datasets to Learn Clean-Data Distribution with Purified Generative Adversarial Networks (AAAI 2023 · better on 2 of 2 shared results)
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
- Learning an Invariant Hilbert Space for Domain Adaptation (CVPR 2017 · similarity 0.92)
- Domain2Vec: Domain Embedding for Unsupervised Domain Adaptation (ECCV 2020 · similarity 0.90)
- Unsupervised Visual Domain Adaptation: A Deep Max-Margin Gaussian Process Approach (CVPR 2019 · similarity 0.90)
- A General Representation-Based Approach to Multi-Source Domain Adaptation (ICML 2025 · similarity 0.89)
- SimProF: A Simple Probabilistic Framework for Unsupervised Domain Adaptation (AAAI 2025 · similarity 0.89)
- Exploring Sparse Visual Prompt for Domain Adaptive Dense Prediction (AAAI 2024 · similarity 0.87)
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.
