Image super‐resolution via dynamic network

Convolutional neural networks depend on deep network architectures to extract accurate information for image super‐resolution.However,obtained information of these con-volutional neural networks cannot completely express predicted high‐quality images for complex scenes.A dynamic network for image super‐resolution(DSRNet)is presented,which contains a residual enhancement block,wide enhancement block,feature refine-ment block and construction block.The residual enhancement block is composed of a residual enhanced architecture to facilitate hierarchical features for image super‐resolution.To enhance robustness of obtained super‐resolution model for complex scenes,a wide enhancement block achieves a dynamic architecture to learn more robust information to enhance applicability of an obtained super‐resolution model for varying scenes.To prevent interference of components in a wide enhancement block,a refine-ment block utilises a stacked architecture to accurately learn obtained features.Also,a residual learning operation is embedded in the refinement block to prevent long‐term dependency problem.Finally,a construction block is responsible for reconstructing high‐quality images.Designed heterogeneous architecture can not only facilitate richer structural information,but also be lightweight,which is suitable for mobile digital devices.Experimental results show that our method is more competitive in terms of performance,recovering time of image super‐resolution and complexity.The code of DSRNet can be obtained at https://github.com/hellloxiaotian/DSRNet.

Dual-Hand Motion Capture by Using Biological Inspiration for Bionic Bimanual Robot Teleoperation

Bionic bimanual robot teleoperation can transfer the grasping and manipulation skills of human dual hands to the bionic bimanual robots to realize natural and flexible manipulation.The motion capture of dual hands plays an important role in the teleoperation.The motion information of dual hands can be captured through the hand detection,localization,and pose estimation and mapped to the bionic bimanual robots to realize the teleoperation.However,although the motion capture technology has achieved great achievements in recent years,visual dual-hand motion capture is still a great challenge.So,this work proposed a dual-hand detection method and a 3-dimensional(3D)hand pose estimation method based on body and hand biological inspiration to achieve convenient and accurate monocular dual-hand motion capture and bionic bimanual robot teleoperation.First,a dual-hand detection method based on body structure constraints is proposed,which uses a parallel structure to combine hand and body relationship features.Second,a 3D hand pose estimation method with bone-constraint loss from single RGB images is proposed.Then,a bionic bimanual robot teleoperation method is designed by using the proposed hand detection and pose estimation methods.Experiment results on public hand datasets show that the performances of the proposed hand detection and 3D hand pose estimation outperform state-of-the-art methods.Experiment results on a bionic bimanual robot teleoperation platform shows the effectiveness of the proposed teleoperation method.