aiqbench analysis
MambaGaze introduces a novel framework using bidirectional Mamba-2 models to address missing eye-tracking data while assessing cognitive load in safety-critical environments like autonomous driving. By explicitly modeling observation uncertainty through XMD encoding, the system achieves state-of-the-art accuracy on standard datasets and demonstrates feasible real-time performance on edge hardware such as NVIDIA Jetson Orin platforms. This research advances human-centered AI by enabling robust vigilance monitoring despite frequent signal interruptions caused by natural blinks or tracking failures.
MambaGaze: Bidirectional Mamba with Explicit Missing Data Modeling for Cognitive Load Assessment from Eye-Gaze Tracking Data
arXiv:2605.22775v3 Announce Type: replace-cross Abstract: Real-time cognitive load assessment from eye-tracking signals could enable adaptive human-centered AI in safety-critical applications such as driver vigilance monitoring or automated flight deck assistance, yet two challenges persist: handling frequent data missingness from blinks and tracking failures, and efficiently modeling long-range temporal dependencies. We propose MambaGaze (Bi-Mamba), a framework that addresses these challenges through (1) XMD encoding, which augments raw features with observation masks and time-deltas to explicitly model data uncertainty, and (2) bidirectional Mamba-2, which captures temporal dependencies with linear computational complexity. Experiments on CLARE and CL-Drive datasets under leave-one-subject-out evaluation show that MambaGaze achieves 77.1% accuracy and 59.2% macro-F1 on CLARE, and 69.4% accuracy and 51.5% macro-F1 on CL-Drive, attaining the highest average LOSO macro-F1 (55.3%) across all ten compared models. Input-stream ablation indicates that log-scaled time-deltas are the strongest single channel in our setting, and combining all three XMD streams provides consistent gains of 5-20 pp macro-F1. Edge deployment benchmarks on three NVIDIA Jetson Orin platforms show real-time inference at 27-36 FPS with power consumption below 6.6 W, supporting feasibility for embedded cognitive load monitoring.
Abstract:Real-time cognitive load assessment from eye-tracking signals could enable adaptive human-centered AI in safety-critical applications such as driver vigilance monitoring or automated flight deck assistance, yet two challenges persist: handling frequent data missingness from blinks and tracking failures, and efficiently modeling long-range temporal dependencies. We propose MambaGaze (Bi-Mamba), a framework that addresses these challenges through (1) XMD encoding, which augments raw features with observation masks and time-deltas to explicitly model data uncertainty, and (2) bidirectional Mamba-2, which captures temporal dependencies with linear computational complexity. Experiments on CLARE and CL-Drive datasets under leave-one-subject-out evaluation show that MambaGaze achieves 77.1% accuracy and 59.2% macro-F1 on CLARE, and 69.4% accuracy and 51.5% macro-F1 on CL-Drive, attaining the highest average LOSO macro-F1 (55.3%) across all ten compared models. Input-stream ablation indicates that log-scaled time-deltas are the strongest single channel in our setting, and combining all three XMD streams provides consistent gains of 5-20 pp macro-F1. Edge deployment benchmarks on three NVIDIA Jetson Orin platforms show real-time inference at 27-36 FPS with power consumption below 6.6 W, supporting feasibility for embedded cognitive load monitoring.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
I.2.6; I.5.4; H.1.2
arXiv:2605.22775 [cs.LG]
(or arXiv:2605.22775v3 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2605.22775 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 280 words
Related coverage
Bidirectional Preference Synthesis: Learning Prompt-Conditioned Preferences from Boundary Failures
arXiv · Oct 7, 2026
GeoNIRS: Geometry-Aware Pretraining for Heterogeneous fNIRS Representation Learning
arXiv · Oct 14, 2026
Bidirectional Semantic Complementary Tool Retrieval for Remote Sensing Agents
arXiv · Jun 9, 2026
Anatomy of Associative Recall in Fixed-State Recurrences: A Matched-State Decomposition, an Interference Wall, and a Curriculum That Breaks It
arXiv · Sep 16, 2026