Robotic arms are widely used in various aspects of human-robot collaboration. The primary goal of this study is to explore the usability of robotic arms for delivering objects to humans in dynamic environments. Traditional robotic arms often face limitations in path planning, such as difficulties adapting to dynamic environments and complex developmental processes. To overcome these challenges, this study employs reinforcement learning (RL) to train four models-the Approach RL Model, Delivery RL Model, Decision RL Model, and Merged Model-as alternatives to conventional path planning control. Typically, there exists a significant discrepancy between simulated data and real-world features. Although image segmentation can substantially reduce the gap between virtual and real environments, notable differences remain in hand features. Therefore, to further bridge the simulation-to-reality gap, this study applies CycleGAN to transform real hand features into virtual hand features, thereby enhancing the model's transferability. Experimental results show that the Decision RL Model achieved an accuracy of 99.17%, while the Merged Model achieved 99.92%. The proposed method effectively improves the stability and accuracy of human-robot collaboration in complex scenarios. Overall, this study validates the feasibility of integrating RL, image segmentation, and image translation techniques, offering a scalable and efficient task-solving solution for robotic arms in highly dynamic application domains.

Premature convergence is a critical issue of Particle Swarm Optimization (PSO). The weak global search capability causes particles trapped in local minima at early stage of learning process. There is several research dedicate to solve this problem over the decade. This paper proposes a new algorithm combined Isolation Forest and Particle Swarm Optimization called Isolation Forest Backward Particle Swarm Optimization (IFB-PSO). The proposed new learning scheme helps particles escaping from local minima. The particle will jump backward to the targeted position when the particle trapped over specific iterations. The destination is precisely selected by Isolation Forest to endow the backward particle hopeful future. IFB-PSO is evaluated by a classic benchmark suite, cart-pole problem, and mountain car problem. Experimental results show that IFB-PSO gets competitive results on the benchmark suite with different dimensions and two control problems in comparison with 11 well-known optimization algorithms. The behavior of backward particles is also analyzed to inspect the utility and efficiency of the backward process.

Human–machine interactions have become increasingly crucial in the current era of the Internet of Things (IoT). Mutual feedback is critical for adjusting machine operations to improve the efficiency of human–machine interactions. Imaging can be easily conducted in various contexts to acquire large volumes of visual information, such as that regarding human gestures. In the present study, machine learning technology, which is the driving technology for intelligent processing in IoT systems, was adopted to develop a system for identifying and classifying six hand gestures and five body poses. Gesture and pose data were collected and analyzed using multiple algorithms to construct classification models for pose recognition. Data for one individual were used to train base gesture and pose recognition models, and transfer learning was then performed to adapt these base models to the gesture and pose data of other individuals. The adapted models achieved satisfactory recognition accuracy. The developed gesture and pose recognition models were tested by employing them to control a robotic arm and an automated guided vehicle, respectively. All models achieved accuracy rates of >97%, thereby confirming the effectiveness of the proposed machine-learning-based method for gesture and pose recognition.

Poststroke injuries limit the daily activities of patients and cause considerable inconvenience. Therefore, predicting the activities of daily living (ADL) results of patients with stroke before hospital discharge can assist clinical workers in formulating more personalized and effective strategies for therapeutic intervention, and prepare hospital discharge plans that suit the patients needs. This study used the leave-one-out cross-validation procedure to evaluate the performance of the machine learning models. In addition, testing methods were used to identify the optimal weak learners, which were then combined to form a stacking model. Subsequently, a hyperparameter optimization algorithm was used to optimize the model hyperparameters. Finally, optimization algorithms were used to analyze each feature, and features of high importance were identified by limiting the number of features to be included in the machine learning models. After various features were fed into the learning models to predict the Barthel index (BI) at discharge, the results indicated that random forest (RF), adaptive boosting (AdaBoost), and multilayer perceptron (MLP) produced suitable results. The most critical prediction factor of this study was the BI at admission. Machine learning models can be used to assist clinical workers in predicting the ADL of patients with stroke at hospital discharge.

In this article, the adaptive-network-based fuzzy inference system (ANFIS) is combined with the double deepQ-network (DDQN) to realize a fuzzy DDQN (FDDQN) such that a humanoid robot can generate a linear inverted pendulum model-based gait pattern in real time. The FDDQN not only allows the humanoid robot to correct the gait pattern instantly but also improves its stability. The proposed scheme is designed and implemented in a toddler-sized humanoid robot called Louis. First, four pressure sensors are installed on the bottom of the sole and one inertial measurement unit is set up on the trunk of the robot. A wireless communication chip is employed to transfer the data to a computer to determine the required parameters for the robot. Next, a control system based on the Linux operating system is developed. The values of the center of pressure and acceleration obtained with the ANFIS are adopted to train the DDQN. The proposed neural network comprises four layers, and the model is cautiously selected to avoid overfitting. The proposed scheme is verified using a robot simulator and then real-time-tested on Louis. The experimental results indicate that the FDDQN can provide the robot timely feedback during walking as well as helps it in adjusting the gait pattern independently. The balancing of the robot through effective dynamic feedback is similar to the balancing ability of an infant learning to walk.

This paper presents a Human Partner System (HPS) that is composed of three sub-systems: a hearing system, a speaking system, and a vision system. Coordinated with these systems, HPS can recognize people and dialog with them. To attract a human's attention for interaction, this paper combines the difference image method and the GrabCut algorithm for HPS. To assist humans in their daily lives and to enhance their interactive fun, HPS has a short-term memory function and can read out new email when a known person appears in her line of sight. To allow HPS to search for a target object in a complicated environment, this paper proposes the double peak continuously adaptive mean shift (DPCAMS) algorithm, which combines the contours and color information of target object. This algorithm is easy to implement and the experimental result demonstrates that this algorithm is faster than the speeded up robust features (SURF) algorithm.

Inverse kinematics is a general method for defining the joint angles of the robot arm. This method provides an efficient way to control the robot arm for several tasks. However, the server motors or the mechanism design of the robot arm may not always be ideal. If the motor consumption is existed, the error of the final position of the robot arm will be increased. In order to solve this problem, this paper proposes an intelligent method for the posture calibration of the robot arm. In this paper, the particle swarm optimization (PSO) algorithm and the proposed neural network model are integrated to calibrate the kinematics of the robot arm. The experimental results show that the control error can be reduced by applying the proposed method. The feasibility and practicality of the proposed method are also validated in the experiments.

This paper presents a recognition system for understanding the words of home-service-related sign language. Because the data received from a sensor are sequential, the hidden Markov model (HMM) that has been successfully applied to speech signals is chosen as a classifier. However, the number of states in the HMM model should be decided upon first before constructing the HMM classifier. To solve this problem, an entropy-based K -means algorithm is proposed to evaluate the number of states in the HMM model with an entropy diagram. Four real datasets are utilized to verify the developed entropy-based K -means algorithm. Moreover, a data-driven method is given to combine the artificial bee colony algorithm with the Baum-Welch algorithm to determine the structure of HMM. The database contains 11 home-service-related Taiwan sign language words and each word is performed ten times, five males and five females are invited to perform such words. Finally, the recognition system is established by 11 HMM models, and the cross-validation demonstrates an average recognition rate of 91.3%.

This paper proposes a control method that improves the ability of adult-sized humanoid robots to adapt to weightlifting situations. In order to achieve the goal of having humanoid robots automatically balance their motion for weight-lifting situations, feedback control is added to the motion control system. The feedback sensors include a three-axis accelerometer and a three-axis gyroscopic, which would be processed by Kalman filter, as well as eight force sensors providing the zero moment point (ZMP) information on the robot. These feedback signals are used as the input of a Fuzzy Q-learning controller, which adjusts the motions to keep the stabilization of the robot. The Fuzzy Q-learning controller consists of two stages: one is the stage of fitting the output weights of each pose in motion patterns, and the second is training the rule-table of the controller. The experiment shows that the controller allows the adult-sized robot to walk stably in weight-lifting situation. Thus, the developed controller indeed keeps the balance of the robot in different situations, which gives the robot the ability to adapt to various environments in the manner of human beings.

The design and implementation of particle swarm optimization (PSO) gait learning method for adult-sized humanoid robots is proposed in this paper. In order to reduce the motor damage and let train motions more convenient, a robotics simulator system for humanoid robots is designed. This robotics simulator system is established by an open source software-Open Dynamics Engine (ODE). The model of David developed by aiRobots laboratory is a combination of rigid bodies and joints. The humanoid robot is trained on the robotics simulator system with PSO method, which chooses the trajectory of robot's center of mass as the fitness value to learn faster and stable gait automatically. The results of the experiment show that the motions which play on the robotics simulator system are very similar to the real motions, so it can be utilized as the motion training platform. The result of the PSO gait learning method has great performance on the robotics simulator system. The humanoid robot learns gait pattern from marking time to moving center of mass and swing its legs. Finally, this gait let the real humanoid robot walk forward at 14.5 cm/s.