Research map / Label-efficient and robust learning / Domain adaptation and generalization
DARE-GRAM: Unsupervised Domain Adaptation Regression by Aligning Inverse Gram Matrices
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 |
|---|---|---|---|
| align inverse-Gram angle and scale | ResNet-18 regression model · dSprites (average across six directions) | MAE: 0.237 -> 0.164 | Helped |
| align inverse-Gram angle and scale | ResNet-18 regression model · MPI3D (average across six directions) | MAE: 0.205 -> 0.160 | Helped |
| align inverse-Gram angle and scale | ResNet-18 regression model · Biwi Kinect (average across two directions) | MAE: 0.280 -> 0.260 | Helped |
| add inverse-Gram angle alignment (ablation) | DARE-GRAM on dSprites C → S · dSprites C → S | MAE: 0.90 -> 0.27 | Helped |
| add inverse-Gram angle alignment (ablation) | DARE-GRAM on dSprites N → S · dSprites N → S | MAE: 0.65 -> 0.36 | Helped |
| add scale alignment (ablation) | DARE-GRAM on dSprites C → S · dSprites C → S | MAE: 0.90 -> 0.23 | Helped |
| add scale alignment (ablation) | DARE-GRAM on dSprites N → S · dSprites N → S | MAE: 0.65 -> 0.60 | Helped |
| combine inverse-Gram angle alignment with scale alignment (ablation) | DARE-GRAM on dSprites C → S · dSprites C → S | MAE: 0.27 -> 0.20 | Helped |
| combine inverse-Gram angle alignment with scale alignment (ablation) | DARE-GRAM on dSprites N → S · dSprites N → S | MAE: 0.36 -> 0.25 | Helped |
| align the Gram matrix rather than its inverse (ablation) | DARE-GRAM on dSprites C → S · dSprites C → S | MAE: 0.27 -> 0.88 | Hurt |
| align the Gram matrix rather than its inverse (ablation) | DARE-GRAM on dSprites N → S · dSprites N → S | MAE: 0.36 -> 0.55 | Hurt |
| align the truncated Gram matrix without inversion (ablation) | DARE-GRAM on dSprites C → S · dSprites C → S | MAE: 0.27 -> 0.89 | Hurt |
Closest papers by meaning
- Co-regularized Alignment for Unsupervised Domain Adaptation (NeurIPS 2018 · similarity 0.89)
- Exploiting Local Feature Patterns for Unsupervised Domain Adaptation (AAAI 2019 · similarity 0.88)
- Pseudo-Labeling for Unsupervised Domain Adaptation with Kernel GLMs (NeurIPS 2026 · similarity 0.88)
- Unsupervised Domain Adaptation Using Feature-Whitening and Consensus Loss (CVPR 2019 · similarity 0.87)
- Minimal-Entropy Correlation Alignment for Unsupervised Deep Domain Adaptation (ICLR 2018 · similarity 0.87)
- CASUAL: Conditional Support Alignment for Domain Adaptation with Label Shift (AAAI 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.