Gaze Estimation on Spresense

TL;DR

This paper presents a lightweight gaze estimation system implemented on the Sony Spresense microcontroller, demonstrating its performance in latency, power consumption, and real-time functionality with a small model running at 3 FPS.

Contribution

The paper introduces TinyTrackerS, a compact gaze estimation model optimized for the Spresense platform, highlighting its architecture and performance metrics.

Findings

TinyTrackerS model size is 169Kb with 85.8k parameters.

Runs at 3 FPS on Spresense with low power consumption.

System achieves acceptable latency and performance for real-time applications.

Abstract

Gaze estimation is a valuable technology with numerous applications in fields such as human-computer interaction, virtual reality, and medicine. This report presents the implementation of a gaze estimation system using the Sony Spresense microcontroller board and explores its performance in latency, MAC/cycle, and power consumption. The report also provides insights into the system's architecture, including the gaze estimation model used. Additionally, a demonstration of the system is presented, showcasing its functionality and performance. Our lightweight model TinyTrackerS is a mere 169Kb in size, using 85.8k parameters and runs on the Spresense platform at 3 FPS.