Growing neural gas (GNG) has been widely used in topological mapping, clustering and unsupervised tasks. It starts from two random nodes and grows until it forms a topological network covering all data. The time required for growth depends on the total amount of data and the current network nodes. To accelerate growth, we introduce a novel distributed batch processing method to extract the rough distribution called Distributed Batch Learning Growing Neural Gas (DBL-GNG). First, instead of using a for loop in standard GNG, we adopt a batch learning approach to accelerate learning. To do this, we replace most of the standard equations with matrix calculations. Next, instead of starting with two random nodes, we start with multiple nodes in different distribution areas. Furthermore, we also propose to add multiple nodes to the network instead of adding them one by one. Finally, we introduce an edge cutting method to reduce unimportant links between nodes to obtain a better cluster network. We demonstrate DBL-GNG on multiple benchmark datasets. From the results, DBL-GNG performs faster than other GNG methods by at least 10 times. We also demonstrate the scalability of DBL-GNG by implementing a multi-scale batch learning process in it, named MS-DBL-GNG, which successfully obtains fast convergence results. In addition, we also demonstrate the dynamic data adaptation of DBL-GNG to 3D point cloud data. It is capable of processing and mapping topological nodes on point cloud objects in real time.

This paper introduces an artificial intelligence (AI) interactive system featuring a self-growing memory network designed to enhance self-efficacy, reduce loneliness, and maintain social interaction among the elderly. The system dynamically analyzes and processes user-written diaries, generating empathic and personalized responses tailored to each individual. The system architecture includes an experience extraction model, a self-growing memory network that provides a contextual understanding of the user’s daily life, a chat agent, and a feedback loop that adaptively learns the user’s behavioral patterns and emotional states. By drawing on both successful and challenging experiences, the system crafts responses that reinforce the self-efficacy of the user, fostering a sense of accomplishment and engagement. This approach improves the psychological well-being of elderly users and promotes their mental health and overall quality of life through consistent interaction. To validate our proposed method, we developed a diary application to facilitate user interaction and collect diary entries. Over time, the system’s capacity to learn and adapt further refines the user experience, suggesting that AI-driven solutions hold significant potential for mitigating the effects of declining self-efficacy on mental health and social interactions. With the proposed system, we achieve an average system usability scale score of 77.3 (SD = 5.4) and a general self-efficacy scale score of 34.2 (SD = 3.5).

Human environments are designed and managed by humans for humans. Thus, adding robots to interact with humans and perform specific tasks appropriately is an essential topic in robotics research. In recent decades, object recognition, human skeletal, and face recognition frameworks have been implemented to support the tasks of robots. However, recognition of activities and interactions between humans and surrounding objects is an ongoing and more challenging problem. Therefore, this study proposed a graph neural network (GNN) approach to directly recognize human activity at home using vision and speech teaching data. Focus was given to the problem of classifying three activities, namely, eating, working, and reading, where these activities were conducted in the same environment. From the experiments, observations, and analyses, this proved to be quite a challenging problem to solve using only traditional convolutional neural networks (CNN) and video datasets. In the proposed method, an activity classification was learned from a 3D detected object corresponding to the human position. Next, human utterances were used to label the activity from the collected human and object 3D positions. The experiment, involving data collection and learning, was demonstrated by using human-robot communication. It was shown that the proposed method had the shortest training time of 100.346 seconds with 6000 positions from the dataset and was able to recognize the three activities more accurately than the deep layer aggregation (DLA) and X3D networks with video datasets.

Mobility support robots can be made available for use by patients in hospital settings. To prevent the robot from being used by unauthorized patients, an identification system must be installed on the robot. The FaceNet system is unable to recognize users when the camera is positioned to view from below, and the true recognition rate is lower when the threshold is set to a higher level. To address these issues, this paper proposes an adaptive face recognition system with a higher threshold setting and the capacity to adapt to bottom camera viewing angles. To achieve high threshold settings, this paper proposes a novel training strategy that enlarges the output distribution range through Negative Feature Augmentation (NFA) when training a Siamese network. The objective of NFA is to introduce additional negative sample pairs during the training phase, thereby enabling the network to expand the recognition distribution range. This hypothesis is demonstrated on the MNIST dataset. Furthermore, the proposed method was tested on the LFW benchmark dataset, resulting in an improved recognition rate of 0.86 when the threshold was set to 0.8. Subsequently, in order to achieve the ability to adaptively recognize users from bottom perspectives, we introduce the use of Growing Neural Gas (GNG) to track feature changes when the user faces different directions while using the robot. Furthermore, we propose methodologies for aligning eyes and standardizing faces, thereby ensuring that the input face is aligned with the learned features. In our experiments, we collected RGB videos of 16 college students using RT-1, and the results demonstrated a high accuracy rate of 0.80, outperforming other methods.

Developing flexible, cost-effective, and durable sensors is a key challenge for integrating Cyber–Physical–Human Systems (CPHSs) into smart homes. This paper introduces a flexible pressure sensor array designed for CPHS applications, addressing the need for cost-effective and durable sensors in smart homes. Our approach combines flexible piezoelectric materials with Swept Frequency Capacitive Sensing (SFCS). Unlike previous pressure sensors made of flexible piezoelectric materials, which can only measure dynamic pressure due to charge leakage, by using SFCS, the piezoelectric material is not directly in the circuit, and our sensor can effectively measure static pressure. While traditional arrays require multiple I/O ports or a matrix configuration, our design measures four distinct locations using only a single I/O port. The sensor is also mechanically flexible and exhibits high durability, capable of functioning even after being cut or torn, provided the electrode contact area remains largely intact. To decode the complex, multiplexed signal from this single channel, we developed a two-stage deep learning pipeline. We utilized data from thin-film resistive pressure sensors as ground truth. A classification model determines which of the four sensors are being touched. Then a regression model uses this touch-state information to estimate the corresponding pressure values. This pipeline employs a hybrid architecture that integrates Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks. The results show that the system can estimate pressure values at each location. To demonstrate its application, the sensor system was integrated into a power recliner, thereby transforming the chair into an interactive tool for daily exercise designed to improve the well-being of older adults. This successful implementation establishes a viable pathway for the development of intelligent, interactive furniture for in-home exercise and rehabilitation within the CPHS paradigm.

Muscle estimation is required to estimate the muscle activity of older adults, and a real-time process can influence the effectiveness of monitoring applications. However, using simulation tools to assess muscle activity consumes much computation energy, which is ineffective for long term monitoring. In this paper, we proposed a novel method named Stacked-LSTM (S-LSTM) multivariate regression model for muscle estimation. S-LSTM contributes to resolving the local information lost in the sequence to-sequence architecture. To validate the proposed model, we recorded human walking on a treadmill using multiple Kinect Azure to form the 3D skeleton joints as the input to our model to predict muscle activity. These skeleton data are passed to Human Musculoskeletal Simulation based on Mujoco Engine to generate the desired outputs for training and testing. Through the experiment, S-LSTM outperforms other multivariate regression models. We measured computation time, and our proposed model runs 0.25 milliseconds faster than the MuJoCo simulation tool.