Three-dimensional (3D) triangulation based on active binocular vision has increasing amounts of applications in computer vision and robotics. An active binocular vision system with non-fixed cameras needs to calibrate the stereo extrinsic parameters online to perform 3D triangulation. However, the accuracy of stereo extrinsic parameters and disparity have a significant impact on 3D triangulation precision. We propose a novel eye gaze based 3D triangulation method that does not use stereo extrinsic parameters directly in order to reduce the impact. Instead, we drive both cameras to gaze at a 3D spatial point P at the optical center through visual servoing. Subsequently, we can obtain the 3D coordinates of P through the intersection of the two optical axes of both cameras. We have performed experiments to compare with previous disparity based work, named the integrated two-pose calibration (ITPC) method, using our robotic bionic eyes. The experiments show that our method achieves comparable results with ITPC.

This paper presents a new type of anthropomorphic robot bionic eyes. Most existing eye mechanisms are serial mechanism and have no degree of freedom (DOF) for eye roll, however, the serial mechanism is not compact enough and the eye roll movement is crucial for image stabilization, consequently, we design a 3 DOFs compact serial-parallel eye mechanism which can implement the eye spherical movements without eccentricity, the diameter of the eyeball is only 45 mm, the total weight of the eye mechanism is 300 g, each of the eye contains two cameras (long and short focal length lens) for mimicking the perception features of human eyes. We also design a 3 DOFs serial neck mechanism for improving the flexibility and the ability of image stabilization of the robot bionic eyes. The neck and each of the eye contain inertial sensors, the robot bionic eyes can implement the vestibuloocular reflex like hybrid image stabilization function in two-stage by using the neck and eye mechanisms. The robot bionic eyes contain completely the same number of DOFs with human which is beneficial to human-robot interaction.

Face attributes prediction has an increasing amount of applications in human–computer interaction, face verification and video surveillance. Various studies show that dependencies exist in face attributes. Multi-task learning architecture can build a synergy among the correlated tasks by parameter sharing in the shared layers. However, the dependencies between the tasks have been ignored in the task-specific layers of most multi-task learning architectures. Thus, how to further boost the performance of individual tasks by using task dependencies among face attributes is quite challenging. In this paper, we propose a multi-task learning using task dependencies architecture for face attributes prediction and evaluate the performance with the tasks of smile and gender prediction. The designed attention modules in task-specific layers of our proposed architecture are used for learning task-dependent disentangled representations. The experimental results demonstrate the effectiveness of our proposed network by comparing with the traditional multi-task learning architecture and the state-of-the-art methods on Faces of the world (FotW) and Labeled faces in the wild-a (LFWA) datasets.