Research map / Detection and segmentation / CNN features
Gaze Target Detection Based on Head-Local-Global Coordination
Yaokun Yang, 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 FOV-based local view alongside global view (ablation) | head-local-global network · GazeFollow | Avg. L2: 0.129 -> 0.114 | Improved |
| add head view features into global-local fusion (ablation) | head-local-global fusion · GazeFollow | Avg. L2: 0.138 -> 0.114 | Improved |
| add global-local position consistency (head attention maps) (ablation) | head-local-global network · GazeFollow | Avg. L2: 0.192 -> 0.114 | Improved |
| add global-local representation consistency (contrastive loss) (ablation) | head-local-global network · GazeFollow | Avg. L2: 0.123 -> 0.114 | Improved |
| add independent heatmap supervision for local and global views (ablation) | head-local-global network · GazeFollow | Avg. L2: 0.126 -> 0.114 | Improved |
| pretrain head branch jointly on Gaze360, ETH-XGaze and MPSGaze (ablation) | head view branch · GazeFollow | Avg. L2: 0.132 -> 0.114 | Improved |
| add local view and coordination to Fang et al. (ablation) | Fang et al. · GazeFollow | Avg. L2: 0.124 -> 0.117 | Improved |
| add local view and coordination to VideoAtt (ablation) | VideoAtt · GazeFollow | Avg. L2: 0.137 -> 0.122 | Improved |
| use local view only instead of global view only (ablation) | head-local-global network · GazeFollow | Avg. L2: 0.129 -> 0.122 | Improved |
| use full method instead of VideoAtt on speed | inference speed · Titan Xp | FPS: 67 -> 45 | Hurt |
Closest papers by meaning
- Object-aware Gaze Target Detection (ICCV 2023 · similarity 0.88)
- End-to-End Human-Gaze-Target Detection With Transformers (CVPR 2022 · similarity 0.88)
- Gaze Target Detection by Merging Human Attention and Activity Cues (AAAI 2024 · similarity 0.88)
- Gaze-LLE: Gaze Target Estimation via Large-Scale Learned Encoders (CVPR 2025 · similarity 0.85)
- From Feature to Gaze: A Generalizable Replacement of Linear Layer for Gaze Estimation (CVPR 2024 · similarity 0.84)
- Toward Semantic Gaze Target Detection (NeurIPS 2024 · similarity 0.83)
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.
