Research map / Label-efficient and robust learning / Recognition and long-tail
EEGEyeNet: a Simultaneous Electroencephalography and Eye-tracking Dataset and Benchmark for Eye Movement Prediction
Ard Kastrati, Martyna Beata Plomecka, Damian Pascual, Lukas Wolf, Victor Gillioz, Roger Wattenhofer, Nicolas Langer
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 deep learning models instead of classical ML | CNN · EEGEyeNet Abs. Position | RMSE (mm): 123 -> 70.2 | Improved |
| Use CNN instead of tree-based models | CNN · EEGEyeNet Angle/Amplitude | angle RMSE (rad): 1.09 -> 0.33 | Improved |
| Use deep models instead of tree-based models on Left-Right | CNN · EEGEyeNet Left-Right | accuracy (%): 97.9 -> 98.3 | Improved |
| Use simpler CNN rather than EEGNet | CNN · EEGEyeNet Abs. Position | RMSE (mm): 81.7 -> 70.2 | Improved |
| Use tree-based models instead of linear models | XGBoost · EEGEyeNet Angle/Amplitude | angle RMSE (rad): 1.39 -> 1.11 | Improved |
| Use RBF SVC/SVR instead of other classical models | RBF SVC/SVR · EEGEyeNet Angle/Amplitude | amplitude RMSE (mm): 74.7 -> 75.9 | Hurt |
| Classical models on absolute position versus naive baseline | KNN · EEGEyeNet Abs. Position | RMSE (mm): 123.3 -> 119.7 | No clear effect |
| Use maximal instead of minimal preprocessing (ablation) | CNN · EEGEyeNet Left-Right | accuracy (%): 98.3 -> 75.5 | Hurt |
| Use maximal instead of minimal preprocessing (ablation) | PyramidalCNN · EEGEyeNet Left-Right | accuracy (%): 98.5 -> 86.6 | Hurt |
| Use maximal instead of minimal preprocessing (ablation) | XGBoost · EEGEyeNet Left-Right | accuracy (%): 97.9 -> 68.3 | Hurt |
| Use maximal instead of minimal preprocessing (ablation) | CNN · EEGEyeNet Abs. Position | RMSE (mm): 70.2 -> 134.6 | Hurt |
| Use maximal instead of minimal preprocessing (ablation) | CNN · EEGEyeNet Angle/Amplitude | angle RMSE (rad): 0.33 -> 1.07 | Hurt |
Closest papers by meaning
- NeuroAtlas: Benchmarking Foundation Models for Clinical EEG and Brain-Computer Interfaces (NeurIPS 2026 · similarity 0.88)
- EEG Benchmarking Needs a Task Specification Layer: NeuroDoc for Rulebook-Guided, Executable Benchmark Construction (NeurIPS 2026 · similarity 0.87)
- EEG2Video: Towards Decoding Dynamic Visual Perception from EEG Signals (NeurIPS 2024 · similarity 0.87)
- BrainOmni: A Brain Foundation Model for Unified EEG and MEG Signals (NeurIPS 2025 · similarity 0.86)
- What do EEG Foundation Models Capture from Human Brain Signals? (NeurIPS 2026 · similarity 0.86)
- EyeGraph: Modularity-aware Spatio Temporal Graph Clustering for Continuous Event-based Eye Tracking (NeurIPS 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.
