Researchers have introduced GMGaze, a novel approach to gaze estimation that utilizes a multi-scale transformer architecture and incorporates context-aware conditioning. This method addresses limitations in existing models by employing early fusion of image features and a Mixture-of-Experts (MoE) design for efficient computational scaling. GMGaze demonstrates state-of-the-art performance on multiple benchmarks, showing improved accuracy in both within-domain and cross-domain gaze estimation tasks. AI
Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →
IMPACT Introduces a new architecture for gaze estimation, potentially improving accuracy and efficiency in applications requiring eye-tracking.
RANK_REASON Academic paper introducing a new model architecture and benchmark results.