Research map / Detection and segmentation / CNN features
Deep Learning-based Pupil Center Detection for Fast and Accurate Eye Tracking System
Kang Il Lee, Jung Ho Jeon, Byung Cheol Song
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 proposed face detection network (FD) (ablation) | Choi2019 pipeline · BioID (e<=0.05) | accuracy: 95.39% -> 95.59% | Improved |
| add proposed glasses removal network (GR) (ablation) | Choi2019 pipeline · BioID (e<=0.05) | accuracy: 95.39% -> 95.79% | Improved |
| add both FD and GR (ablation) | Choi2019 pipeline · BioID (e<=0.05) | accuracy: 95.39% -> 95.99% | Improved |
| add MI representation learning segmentation network (ablation) | Choi2019 pipeline with FD and GR · BioID (e<=0.05) | accuracy: 95.99% -> 96.71% | Improved |
| replace Choi2019 glasses removal network with lightweight proposed one | glasses removal network · Complexity | FLOPs: 245G -> 59G | Improved |
| replace Choi2019 glasses removal network with proposed one | glasses removal network · Complexity | latency: 10ms -> 4ms | Improved |
| use proposed face detection network instead of Faceboxes | face detection network · Complexity | latency: 10ms -> 7ms | Improved |
| use proposed face detection network instead of Faceboxes | face detection network · Complexity | FLOPs: 0.09G -> 0.15G | Hurt |
| proposed system versus Levin2018 | PCT system · BioID (e<=0.1) | accuracy: 99.52% -> 98.95% | Hurt |
| proposed system versus Choi2019 | PCT system · GI4E (e<=0.1) | accuracy: 99.84% -> 99.84% | No clear effect |
| proposed system versus Levin2018 | PCT system · Talking Face Video (e<=0.05) | accuracy: 95.62% -> 96.38% | Improved |
| add perceptual loss with PCCL to glasses removal (ablation) | glasses removal network · customized glasses dataset (qualitative) | image sharpness | Improved |
Closest papers by meaning
- Automatic Gaze Analysis: A Survey of Deep Learning Based Approaches (TPAMI 2023 · similarity 0.83)
- Eye Tracking for Everyone (CVPR 2016 · similarity 0.83)
- A Self Validation Network for Object-Level Human Attention Estimation (NeurIPS 2019 · similarity 0.83)
- E-Gaze: Gaze Estimation With Event Camera (TPAMI 2024 · similarity 0.82)
- Rapidly Deploying On-Device Eye Tracking by Distilling Visual Foundation Models (ECCV 2026 · similarity 0.82)
- EyeGraph: Modularity-aware Spatio Temporal Graph Clustering for Continuous Event-based Eye Tracking (NeurIPS 2024 · similarity 0.81)
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.
