The application of motion primitives to encode robot motion has garnered considerable attention in the field of academic research. Existing models predominantly focus on reproducing task trajectory in relation to position, often neglecting the significance of orientation. Orientation Probabilistic Movement Primitives (ProMPs) indirectly encode motion primitives for attitude by utilizing their trajectory probabilities on Riemannian manifolds, specifically the 3-sphere S3. However, assuming a Gaussian distribution imposes constraints on its abilities. We propose Mixed Orientation ProMPs to enhance trajectory planning and minimize the occurrence of singular configurations. This model consists of multiple separate Gaussian distributions in the tangent space, enabling the approximation of any distribution. Furthermore, optimization objective functions of the Lagrangian type can incorporate constraints, such as singularity avoidance, and others. Finally, the effectiveness and reliability of the algorithm were validated through trajectory planning experiments conducted on the UR5 robotic arm. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

The problem of Person Re-Identification is still a big challenge, as the complex network structure and unsatisfactory generalization performance of widely used deep neural networks make them unsuitable for application to real-world problems. In this paper, we propose a global feature-based person re-identification network with strong generalization. The extracted features part contains two channels of feature fusion: the feature extraction module and the feature generalization module. The feature generalization module is a new MixStyle module added to the feature extraction module, which can effectively mix the style information of images under different domains or even the same domain to form multiple potential domain features, thus improving the generalization performance of the model. In addition, this paper also makes some improvements to the loss function by adding a new constraint on the positive sample pair distance, which makes it possible to maximizes the reduction of intra-class distance in addition to pushing the distance between different classes during the training process. Experimental results on two datasets, Market1501 and DukeMTMC, demonstrate that the method proposed in this paper exhibits strong generalization performance for the person re-identification problem and outperforms current global feature-based person re-identification methods. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

Robot vision, which integrates measurement and perception, plays an important role in robot manipulation applications. Although the current deep learning-based visual neural network models have high perception capabilities, their network structures are usually too large to be implemented on embedded devices for robot vision. In this paper, we propose neural network architecture search (NAS) that combines pre-training and pruning operations to simplify deep neural network architectures. It not only solves this problem without losing network accuracy, but also significantly alleviates the difficulties of long network computation time and redundant search space in traditional NAS methods. Finally, the experimental results show that the neural network generated by the proposed algorithm outperforms the artificially designed neural network, which demonstrates the effectiveness of the method. At the end of the paper, the rationality of the method is proved by experiments and comparisons. The performance of the new algorithm and the generated neural network is better than that of the artificially designed neural network. © 2023 IEEE.

Guided policy search methods (GPSs) have become important methods in the field of reinforcement learning in recent years. GPSs are a kind of policy search methods that utilize trajectory optimization methods to generate training data, guiding supervised learning. In theoretical research, GPSs combine convex optimization and deep learning, and have achieved fruitful results. In practical applications, they have achieved good results in complex control fields such as robots learning, especially manipulator operations. This paper mainly elaborates the development process and improvement route of GPSs. Firstly, the theoretical knowledge related to the GPSs is introduced. Secondly, the framework and basic methods of GPSs are analyzed; Thirdly, various improved GPSs based on the basic methods are generalized. Finally, the development and future improvement directions of GPSs are summarized, and the problems and future development trends are discussed.
© Published under licence by IOP Publishing Ltd.

Guided policy search methods (GPSs) have become important methods in the field of reinforcement learning in recent years. GPSs are a kind of policy search methods that utilize trajectory optimization methods to generate training data, guiding supervised learning. In theoretical research, GPSs combine convex optimization and deep learning, and have achieved fruitful results. In practical applications, they have achieved good results in complex control fields such as robots learning, especially manipulator operations. This paper mainly elaborates the development process and improvement route of GPSs. Firstly, the theoretical knowledge related to the GPSs is introduced. Secondly, the framework and basic methods of GPSs are analyzed; Thirdly, various improved GPSs based on the basic methods are generalized. Finally, the development and future improvement directions of GPSs are summarized, and the problems and future development trends are discussed.
© Published under licence by IOP Publishing Ltd.