EM-DARTS: Hierarchical Differentiable Architecture Search for Eye Movement Recognition

2024-09-22

arXiv.org

AI Summary

Existing eye-tracking recognition methods suffer from heavy reliance on handcrafted architectures and insufficient structural diversity in DARTS-based neural architecture search (NAS), further exacerbated by depth inconsistency between search and evaluation phases. To address these issues, we propose Hierarchical Differentiable Architecture Search (H-DARTS), which jointly optimizes both the global network topology and cell-level structures within a supernet via alternating global-local optimization. We introduce Transfer Entropy for the first time to quantify inter-layer information redundancy, enabling principled search space pruning and architecture compression. This marks the first application of transfer entropy to redundancy assessment in eye-tracking models and resolves the depth-inconsistency limitation inherent in standard DARTS. Evaluated on GazeBase, JuDo1000, and EMglasses, our method achieves state-of-the-art equal error rates (EER) of 0.0453, 0.0377, and 0.1385, respectively—significantly outperforming prior approaches.

Technology Category

• Search and Optimization: Heuristic Search
• Machine Learning: Hardware-aware ML
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Security and Privacy: Large-scale security measurements

Abstract

Eye movement biometrics has received increasing attention thanks to its highly secure identification. Although deep learning (DL) models have shown success in eye movement recognition, their architectures largely rely on human prior knowledge. Differentiable Neural Architecture Search (DARTS) automates the manual process of architecture design with high search efficiency. However, DARTS typically stacks multiple cells to form a convolutional network, which limits the diversity of architecture. Furthermore, DARTS generally searches for architectures using shallower networks than those used in the evaluation, creating a significant disparity in architecture depth between the search and evaluation phases. To address this issue, we propose EM-DARTS, a hierarchical differentiable architecture search algorithm to automatically design the DL architecture for eye movement recognition. First, we define a supernet and propose a global and local alternate Neural Architecture Search method to search the optimal architecture alternately with a differentiable neural architecture search. The local search strategy aims to find an optimal architecture for different cells while the global search strategy is responsible for optimizing the architecture of the target network. To minimize redundancy, transfer entropy is proposed to compute the information amount of each layer, thereby further simplifying the network search process. Experimental results on three public datasets demonstrate that the proposed EM-DARTS is capable of producing an optimal architecture that leads to state-of-the-art recognition performance, {Specifically, the recognition models developed using EM-DARTS achieved the lowest EERs of 0.0453 on the GazeBase dataset, 0.0377 on the JuDo1000 dataset, and 0.1385 on the EMglasses dataset.

Problem

• Eye Movement Recognition
• Deep Learning Models
• Automated Model Design

Innovation

• EM-DARTS
• Transfer Entropy
• Gaze Recognition Optimization