Research map / Learning theory and algorithms / Sample complexity
Scanpath Complexity: Modeling Reading Effort Using Gaze Information
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 |
|---|---|---|---|
| combine fixation and saccade attributes in a supervised regression (ScaComp L) instead of total fixation duration alone | scanpath complexity · Word Count correlation | Spearman correlation: 0.84 -> 0.94 | Helped |
| use heuristic product ScaComp H instead of total fixation duration | scanpath complexity · linguistic complexity measures (Table 3) | Spearman correlation: W 0.84 -> 0.92; CN/C 0.69 -> 0.65 | Mixed |
| use ScaComp L instead of total fixation duration on lexical measures (LD, LS, CN/C) | scanpath complexity · lexical density | Spearman correlation: 0.30 -> 0.22 | Hurt |
| add saccadic attributes to fixational ones in the regression (ablation) | ScaComp L · 10-fold CV on gaze dataset | mean absolute error: 0.1938 -> 0.1912 | Helped |
| add the NLL saccade-likelihood attribute (ablation) | ScaComp L · correlation with linguistic measures | Spearman correlation | Helped |
| add FD and RC attributes (ablation) | ScaComp L · correlation with linguistic measures | Spearman correlation | Helped |
| use all attributes instead of only FD, RC and NLL (ablation) | ScaComp L · correlation with linguistic measures | Spearman correlation | No clear effect |
Closest papers by meaning
- ST-DiffEye: Diffusion-based Continuous Gaze Generation via Joint Scanpath-Trajectory Modeling (NeurIPS 2026 · similarity 0.82)
- EyEar: Learning Audio Synchronized Human Gaze Trajectory Based on Physics-Informed Dynamics (AAAI 2025 · similarity 0.81)
- Beyond Average: Individualized Visual Scanpath Prediction (CVPR 2024 · similarity 0.80)
- Self-Supervised Collaborative Information Bottleneck for Text Readability Assessment (AAAI 2025 · similarity 0.80)
- Magnifying What Matters: Attention-Guided Adaptive Rendering for Visual Text Comprehension (NeurIPS 2026 · similarity 0.80)
- Do Vision-Language Models Measure Up? Benchmarking Visual Measurement Reading with MeasureBench (CVPR 2026 · similarity 0.80)
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.