3ET: Efficient Event-based Eye Tracking using a Change-Based ConvLSTM Network

TL;DR

This paper introduces a novel sparse CB-ConvLSTM model for event-based eye tracking, significantly improving efficiency and accuracy for real-time applications in wearable AR/VR devices by leveraging retina-inspired event cameras.

Contribution

The paper proposes a new CB-ConvLSTM architecture with delta-encoded recurrent paths that reduces computational load while maintaining high accuracy in event-based pupil tracking.

Findings

Achieves approximately 4.7× reduction in arithmetic operations.

Outperforms conventional CNN structures in accuracy.

Enables real-time eye tracking on resource-constrained devices.

Abstract

This paper presents a sparse Change-Based Convolutional Long Short-Term Memory (CB-ConvLSTM) model for event-based eye tracking, key for next-generation wearable healthcare technology such as AR/VR headsets. We leverage the benefits of retina-inspired event cameras, namely their low-latency response and sparse output event stream, over traditional frame-based cameras. Our CB-ConvLSTM architecture efficiently extracts spatio-temporal features for pupil tracking from the event stream, outperforming conventional CNN structures. Utilizing a delta-encoded recurrent path enhancing activation sparsity, CB-ConvLSTM reduces arithmetic operations by approximately 4.7$\times$ without losing accuracy when tested on a \texttt{v2e}-generated event dataset of labeled pupils. This increase in efficiency makes it ideal for real-time eye tracking in resource-constrained devices. The project code and dataset are openly available at \url{https://github.com/qinche106/cb-convlstm-eyetracking}.