aiqbench analysis
EyeTAG introduces a novel causal multi-frame framework that explicitly models gaze trajectory as a kinematic token to eliminate systematic saccade bias in head-eye motion scenarios. By differentiating recent predictions, the method achieves subject-invariant motion representation, reducing mean angular error on standard benchmarks without relying solely on implicit temporal context. This advancement directly addresses a core limitation in current computer vision architectures for dynamic human-computer interaction and driver monitoring systems.
EyeTAG: Eye Trajectory-Aware Gaze Estimation
arXiv:2610.00922v1 Announce Type: cross Abstract: Gaze estimation under natural head-eye motion underpins applications from driver monitoring to human-computer interaction. Single-frame methods predict each frame independently, so consecutive outputs fluctuate as jitter. Multi-frame methods reduce this, but they learn motion implicitly inside appearance features, so the gaze trajectory is never an explicit variable. We propose EyeTAG (Eye Trajectory-Aware Gaze Estimation), a causal multi-frame framework built around an explicit first-order gaze prior: at each step it differentiates its own recent predictions and feeds the resulting trajectory back as a compact kinematic token. Because differencing is translation-invariant in gaze space, this token carries subject-invariant motion rather than personal gaze offsets. Face and eye streams supply visual evidence, fused by cross-attention and a causal Transformer decoder. EyeTAG reduces the mean angular error by about 1.0$^\circ$ on Gaze360 and performs on par with the strongest baseline on EVE (2.56$^\circ$ vs. 2.58$^\circ$). Within-model ablations, which keep the encoder and the rest of the architecture fixed and vary only the gaze history, show that the differential formulation, rather than temporal context alone, removes the systematic saccade bias that persists even with an absolute gaze-history prior. Our code is available at https://github.com/peter8366/EyeTAG.
Abstract:Gaze estimation under natural head-eye motion underpins applications from driver monitoring to human-computer interaction. Single-frame methods predict each frame independently, so consecutive outputs fluctuate as jitter. Multi-frame methods reduce this, but they learn motion implicitly inside appearance features, so the gaze trajectory is never an explicit variable. We propose EyeTAG (Eye Trajectory-Aware Gaze Estimation), a causal multi-frame framework built around an explicit first-order gaze prior: at each step it differentiates its own recent predictions and feeds the resulting trajectory back as a compact kinematic token. Because differencing is translation-invariant in gaze space, this token carries subject-invariant motion rather than personal gaze offsets. Face and eye streams supply visual evidence, fused by cross-attention and a causal Transformer decoder. EyeTAG reduces the mean angular error by about 1.0$^\circ$ on Gaze360 and performs on par with the strongest baseline on EVE (2.56$^\circ$ vs. 2.58$^\circ$). Within-model ablations, which keep the encoder and the rest of the architecture fixed and vary only the gaze history, show that the differential formulation, rather than temporal context alone, removes the systematic saccade bias that persists even with an absolute gaze-history prior. Our code is available at this https URL.
Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2610.00922 [cs.CV]
(or arXiv:2610.00922v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2610.00922 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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 · 291 words
Related coverage
"About a quarter of them contained bugs": Gemini 4's ML R&D benchmark
LessWrong · Oct 16, 2026
Preserving Mathematical Reasoning in Compressed Diffusion Language Models via Trajectory-Aware Low-Rank Approximation
arXiv · Oct 6, 2026
GLM-5.3 Sometimes Considers How Train-Time Actions Shape Future Behavior
LessWrong · Oct 15, 2026
SAP-CoPE: Social-Aware Planning using Cooperative Pose Estimation with Infrastructure Sensor Nodes
arXiv · Oct 12, 2026