2D Image head pose estimation via latent space regression under occlusion settings. (May 2023)

• Record Type: Journal Article
• Authors: Celestino, José; Marques, Manuel; Nascimento, Jacinto C.; Costeira, João Paulo
• Is Part Of: Pattern recognition. Volume 137(2023)
• Publication Date: 2023-05
• Subjects: Head pose estimation -- Occlusion -- Latent space -- Euler angles; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2022.109288
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: Occlusion robust HPE framework relying on latent space regression with multi-loss. Results surpass state-of-the-art methods in face images with occlusion and are similar without occlusion. Ablation study on occlusion size reveals 39% error decrease for large occlusions. Application in real scenario: feeding robot for people with upper body impairment. Synthetic occlusion generation and application to existing head pose datasets. Abstract: Head orientation is a challenging Computer Vision problem that has been extensively researched having a wide variety of applications. However, current state-of-the-art systems still underperform in the presence of occlusions and are unreliable for many task applications in such scenarios. This work proposes a novel deep learning approach for the problem of head pose estimation under occlusions. The strategy is based on latent space regression as a fundamental key to better structure the problem for occluded scenarios. Our model surpasses several state-of-the-art methodologies for occluded HPE, and achieves similar accuracy for non-occluded scenarios. We demonstrate the usefulness of the proposed approach with: (i) two synthetically occluded versions of the BIWI and AFLW2000 datasets, (ii) real-life occlusions of the Pandora dataset, and (iii) a real-life application to human-robot interaction scenarios where face occlusions often occur. Specifically, the autonomous feeding from a robotic arm.