Research map / Label-efficient and robust learning / Domain adaptation and generalization
Through the Frequency Lens: Cross-Domain Generalisable Gaze Estimation with Adaptive Modulation
Yang Xu, Yiwei Bao, Feng Lu
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 |
|---|---|---|---|
| add AISM (adaptive interference suppression) (ablation) | ResNet-18 · cross-domain gaze average | angular error: 8.54 -> 6.69 | Improved |
| add SDM (spectrum diversification) (ablation) | ResNet-18 · cross-domain gaze average | angular error: 8.54 -> 6.96 | Improved |
| add SDM on top of AISM (ablation) | ResNet-18 · cross-domain gaze average | angular error: 6.69 -> 6.55 | Improved |
| add SDM on top of AISM (ablation) | ResNet-50 · DG->DM | angular error: 6.45 -> 5.85 | No clear effect |
| use adaptive AISM instead of fixed band suppression (ablation) | ResNet-18 · cross-domain gaze average | angular error: 7.79 -> 6.69 | Improved |
| suppress lowest-frequency band 1 for all images (ablation) | ResNet-18 · cross-domain gaze average | angular error: 8.54 -> 13.03 | Hurt |
| apply modulation maps at inference (ablation) | ResNet-18 with AISM · cross-domain gaze average | angular error: 6.79 -> 6.69 | Improved |
| train with AISM but do not modulate at inference (ablation) | ResNet-18 · cross-domain gaze average | angular error: 8.54 -> 6.79 | Improved |
| use SDM frequency diversification instead of spatial augmentation (ablation) | ResNet-18 · cross-domain gaze average | angular error: 7.75 -> 6.96 | Improved |
| spatial augmentation (color jitter and affine) (ablation) | ResNet-50 · DE->DM | angular error: 6.84 -> 7.06 | Hurt |
| swap ResNet-18 for ResNet-50 baseline | gaze estimation baseline · cross-domain gaze average | angular error: 8.54 -> 8.08 | Improved |
Closest papers by meaning
- Escaping the Low-Frequency Bias: Adversarial Frequency Perturbation for Generalisable Gaze Estimation (ECCV 2026 · similarity 0.91)
- Learning a Generalized Gaze Estimator from Gaze-Consistent Feature (AAAI 2023 · similarity 0.89)
- Gaze from Origin: Learning for Generalized Gaze Estimation by Embedding the Gaze Frontalization Process (AAAI 2024 · similarity 0.88)
- From Feature to Gaze: A Generalizable Replacement of Linear Layer for Gaze Estimation (CVPR 2024 · similarity 0.88)
- A Generalized Label Shift Perspective for Cross-Domain Gaze Estimation (NeurIPS 2025 · similarity 0.88)
- Gaze Label Alignment: Alleviating Domain Shift for Gaze Estimation (AAAI 2025 · 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.
