The field of computer vision is moving towards more accurate and efficient methods for gaze estimation and object pose estimation. Recent developments have focused on improving the robustness and generalizability of these methods, with a particular emphasis on real-time applications and edge AI solutions. Noteworthy papers in this area include RTGaze, which achieves state-of-the-art performance in gaze redirection, and CoordAR, which presents a novel autoregressive framework for one-reference 6D pose estimation of unseen objects. Other notable works include OPFormer, which integrates object detection and pose estimation with a versatile onboarding process, and WALDO, which proposes a dynamic non-uniform dense sampling strategy for model-based 6D pose estimation under occlusion. These advancements have significant implications for applications such as robotics, augmented reality, and scene understanding.
Advances in Gaze Estimation and Object Pose Estimation
Sources
RTGaze: Real-Time 3D-Aware Gaze Redirection from a Single Image
Toward Gaze Target Detection of Young Autistic Children
6D Strawberry Pose Estimation: Real-time and Edge AI Solutions Using Purely Synthetic Training Data
3D Gaussian and Diffusion-Based Gaze Redirection
Do Blind Spots Matter for Word-Referent Mapping? A Computational Study with Infant Egocentric Video
OPFormer: Object Pose Estimation leveraging foundation model with geometric encoding
CoordAR: One-Reference 6D Pose Estimation of Novel Objects via Autoregressive Coordinate Map Generation
Hybrid-Domain Adaptative Representation Learning for Gaze Estimation
End-to-End Multi-Person Pose Estimation with Pose-Aware Video Transformer
Learning to See Through a Baby's Eyes: Early Visual Diets Enable Robust Visual Intelligence in Humans and Machines
Computer Vision Modeling of the Development of Geometric and Numerical Concepts in Humans
WALDO: Where Unseen Model-based 6D Pose Estimation Meets Occlusion
Box6D : Zero-shot Category-level 6D Pose Estimation of Warehouse Boxes
Physics-Informed Machine Learning for Efficient Sim-to-Real Data Augmentation in Micro-Object Pose Estimation