Pupil.IO AIO Logo Pupil.IO AIO

Pupil.IO AIO
High-Speed AI Eye-Tracking System

Research-grade high-precision eye movement & pupil data for brain science, mental health assessment and commercial transformation.

200/400/2000Hz

Sampling Rate

0.05°

Spatial Accuracy

82%

MCI Accuracy

3-5mins

Rapid Testing

Pupil.IO AIO Device

About Pupil.IO AIO

Pupil.IO AIO Applications

Pupil.IO AIO is a high-performance eye-tracking system for medical and scientific research applications developed by Hangzhou DeepGaze Technology Co., Ltd. It delivers research-grade high-precision eye movement and pupil data, and provides a complete solution from data collection to analysis and application for brain science research, mental health assessment and related commercial transformation.

Clinical-Grade Precision

Meets all medical application needs with high-precision data

Easy Operation

2-point calibration, no head support required

All-Age Adaptation

Robust algorithm for low-cooperation users of all ages

Typical Applications

Pupil.IO AIO provides precise eye-tracking solutions for various mental health and medical research scenarios

Community MCI Rapid Screening

Community MCI Rapid Screening

3-4 minutes of eye-tracking task for MCI detection, 82% accuracy with excellent sensitivity and specificity. Well received by elderly users in community deployment.

Algorithm Accuracy AUC-ROC Sensitivity Specificity
Logistic Regression 82% 0.91 80% 84%
Random Forest 80% 0.89 78% 82%
Ward Intelligent Robot

Ward Intelligent Robot

Built-in Pupil.IO AIO eye-tracking module in BuddyBot (the first domestic ward robot based on eye-tracking + LLM). Launched in Xuanwu Hospital and JD Elderly Care Platform, automatically completes cognitive & mental health monitoring.

Xuanwu Hospital of Capital Medical University
ASD Early Screening

ASD Early Screening

AI-based ASD screening system developed with Hangzhou Institute of Medical Sciences, Chinese Academy of Sciences. 3-5 minutes for rapid screening with >83% sensitivity and specificity.

Served nearly 10,000 children in health system pilot
Depression Risk Assessment

Depression Risk Assessment

Accuracy of 75-85% for college student depression risk screening. Deployed in psychological empowerment centers, providing objective physiological indicators for adolescent depression prevention.

Objective physiological indicators for precise assessment

Applicable Fields

Autism Screening Eye-Controlled Interaction Cognitive Physical Examination Reading Disorder Diagnosis Depression Diagnosis Neurodegenerative Disease Screening

Technical Advantages

AI Neural Network Algorithm

AI Neural Network Algorithm

More robust, high tolerance for low-cooperation users

Simplified Calibration

2-point calibration only, no head support required, easy to use

High-Speed Sampling

200/400/2000Hz high-precision eye movement & pupil data

System-Level Solutions

OEM/ODM solutions for brain science and mental health applications

All-Age Adaptation

Suitable for users of all age groups (children/adolescents/elderly)

Technical Specifications

Parameters Commercial Research Unit Notes
Sampling Rate 200Hz 400/2000Hz Hz Adapt to diverse needs
Spatial Accuracy ≤0.1° ≤0.05° ° Research-grade precision
Latency ≤8ms ≤5ms ms Real-time data output
Working Distance 50-70cm 50-70cm cm No head support needed
Calibration 2-point 2-point - One-click operation
Software Development Kit C++/Python/Matlab C++/Python/Matlab - Lifetime free use

Note: Pre-release data (2026Q2-2026Q4), all test data based on official laboratory verification

Sampling Rate Chart

High-Precision Spatiotemporal Analysis

200Hz high-speed sampling & 0.05° spatial accuracy for precise spatiotemporal analysis of saccade and other eye movement patterns (meets rigorous commercial and research needs).

3D Gaze Data Output

Complete Binocular 3D Gaze Data

Output full binocular 3D Gaze data (screen coordinates, pupil 3D position, gaze direction vector, pupil diameter, etc.) — a complete data foundation for ophthalmic clinical applications.

ID-T Algorithm for Automatic Event Extraction

Built-in ID-T algorithm, one line of code to automatically extract fixation, saccade, blink and other key eye movement events (eliminate tedious underlying data processing).

Pupil.IO Previewer

Multi-Platform Support & Development Tools

  • Pupil.IO Previewer (Android/Windows)
  • Native PsychoPy support
  • Real-time facial image output (up to 200 FPS)
  • Lifetime free SDK with free upgrades

Partners & Cases

Scientific Research Partners

Beijing University of Posts and Telecommunications Hangzhou Institute of Medical Sciences Lanzhou University Hangzhou Normal University

Enterprise Partners

Brain Aurora Technology Virtual Reality & Reality Technology

Contact Us

Have questions or want to request a demo? Fill out the form below and we'll get back to you shortly.

Company Information

Company

Hangzhou DeepGaze Technology Co., Ltd.

Address

Room 211, Building B, Aili Center, Cangqian Street, Yuhang District, Hangzhou, Zhejiang Province, China

Send Us an Inquiry