Researchers have developed a new method to improve the accuracy of geometric eye trackers, which are crucial for gaze-based interaction and multimodal studies. The proposed technique involves a lightweight neural refiner that combines predictions from multiple calibrators to reduce session-specific calibration errors. When tested on a controlled dataset, this approach significantly lowered the mean angular error compared to classical methods, demonstrating its potential for enhancing the reliability of gaze as a behavioral signal in interactive modeling. AI
IMPACT Enhances the reliability of gaze-based interaction and multimodal studies by improving eye-tracking accuracy.
RANK_REASON The item is a research paper published on arXiv detailing a new method for improving eye-tracking accuracy. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Geometric eye trackers
- Gotit.pub
- Hugging Face
- neural refiner
- ScienceCast
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →