Research map / Graph learning / Knowledge and temporal graphs
Beyond Scanpaths: Graph-Based Gaze Simulation in Dynamic Scenes
Luke Palmer, Petar Palasek, Hazem Abdelkawy
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 |
|---|---|---|---|
| use ART instead of HEAT (ablation) | ART+ODN · Focus100 test | TC: 0.13 -> 0.22 | Improved |
| use ART instead of HEAT (ablation) | ART+ODN · Focus100 test | DTW: 59.50 -> 42.31 | Improved |
| use ART instead of HGT (ablation) | ART+ODN · Focus100 test | LEV: 1.28 -> 1.23 | Improved |
| use ART instead of HGT (ablation) | ART+ODN · Focus100 test | DTW: 42.72 -> 42.31 | No clear effect |
| use ODN head instead of MDN with k=10 (ablation) | ART · Focus100 test | TC: 0.14 -> 0.22 | Improved |
| use ODN head instead of MDN with k=20 (ablation) | ART · Focus100 test | DTW: 45.69 -> 42.31 | Improved |
| increase graph history from 8 to 20 timesteps (ablation) | ART+ODN · Focus100 test | TC: 0.17 -> 0.22 | Improved |
| increase graph history from 1 to 20 timesteps (ablation) | ART+ODN · Focus100 test | DTW: 42.35 -> 42.31 | No clear effect |
| ART vs strongest saliency baselines on saliency maps | ART · Focus100 test | NSS: 4.481 -> 4.864 | Improved |
| ART vs ViNet on MAAD saliency | ART · MAAD | NSS: 5.733 -> 4.926 | Hurt |
| ART vs baselines on fixation rate | ART · Focus100 test | Fix Rate: 0.57 -> 1.64 | Improved |
Closest papers by meaning
- ST-DiffEye: Diffusion-based Continuous Gaze Generation via Joint Scanpath-Trajectory Modeling (NeurIPS 2026 · similarity 0.89)
- FIFA: Fine-grained Inter-frame Attention for Driver's Video Gaze Estimation (CVPR 2025 · similarity 0.88)
- DiffEye: Diffusion-Based Continuous Eye-Tracking Data Generation Conditioned on Natural Images (NeurIPS 2025 · similarity 0.86)
- GazeFlow: From Human Gaze Behavior to Generative Egocentric Gaze Prediction (NeurIPS 2026 · similarity 0.86)
- Unifying Top-down and Bottom-up Scanpath Prediction Using Transformers (CVPR 2024 · similarity 0.86)
- What Do You See in Vehicle? Comprehensive Vision Solution for In-Vehicle Gaze Estimation (CVPR 2024 · similarity 0.85)
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.
