Research map / Label-efficient and robust learning / Recognition and long-tail
Controllable Invariance through Adversarial Feature Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use adversarial feature learning to remove sensitive information | Controllable Invariance · German | accuracy: 0.744 vs 0.727 | Helped |
| use adversarial feature learning to remove sensitive information | Controllable Invariance · Adult | accuracy: 0.844 vs 0.813 | Helped |
| use an invariant representation | Controllable Invariance · Adult biased categories | accuracy: 0.788 vs 0.748 | Helped |
| use an invariant representation | Controllable Invariance · German biased categories | accuracy: 0.676 vs 0.648 | Helped |
| use adversarial invariant feature learning | Controllable Invariance · IWSLT 2015 fr-en test | BLEU: +0.6 BLEU | Helped |
| use adversarial invariant feature learning | Controllable Invariance · IWSLT 2015 de-en test | BLEU: 28.1 vs 27.7 | Helped |
| remove the discriminator (ablation) | Controllable Invariance · IWSLT 2015 fr-en test | BLEU: 36.1 -> 35.3 | Hurt |
| remove the discriminator (ablation) | Controllable Invariance · IWSLT 2015 de-en test | BLEU: 28.1 -> 27.6 | Hurt |
| use a shared encoder instead of separate language encoders (ablation) | Controllable Invariance · IWSLT 2015 fr-en test | BLEU: 36.1 -> 35.4 | Hurt |
| use a shared encoder instead of separate language encoders (ablation) | Controllable Invariance · IWSLT 2015 de-en test | BLEU: 28.1 -> 27.7 | Hurt |
| use adversarial invariant feature learning | Controllable Invariance · Extended Yale B | identity classification accuracy: 0.89 vs 0.85 | Helped |
| use adversarial invariant feature learning | Controllable Invariance · Extended Yale B | lighting classification accuracy: 0.96 -> 0.57 | Helped |
Closest papers by meaning
- Invariant Representations without Adversarial Training (NeurIPS 2018 · similarity 0.93)
- A representation-learning game for classes of prediction tasks (ICLR 2024 · similarity 0.87)
- Invariant Representations through Adversarial Forgetting (AAAI 2020 · similarity 0.87)
- Unsupervised Adversarial Invariance (NeurIPS 2018 · similarity 0.86)
- Learning Adversarially Robust Representations via Worst-Case Mutual Information Maximization (ICML 2020 · similarity 0.86)
- Rethinking Invariance Regularization in Adversarial Training to Improve Robustness-Accuracy Trade-off (ICLR 2025 · similarity 0.86)
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.