Virtual reality (VR) is increasingly capable and inexpensive, and VR devices have become indispensable in many domains, such as gaming, videoconferencing, education, and healthcare. VR has also been applied to music performance and learning. Virtual instruments, such as virtual pianos and drums, free users from the need to own physical forms of these (often bulky and expensive) instruments. VR devices enable users to enjoy music anytime and anywhere without constraints. Virtual concerts, including spatial audio simulations and reconstructions of historical performances, are becoming increasingly common. Previous studies have primarily examined virtual guitars in non-VR environments and air guitar chord recognition. However, systematic research on virtual air guitar systems in VR remains scarce. Virtual guitar games that are available on the market cannot recognize hand gestures accurately and thus cannot accurately identify the strumming patterns and chords played by the player. To overcome this problem, we propose a VR-based virtual air guitar system that can recognize 30 chords and various strumming techniques through deep learning and visual feedback. Employing a black-box approach, we combine WaveNet and FiLM to simulate electric guitar pedal effects with a knob difference loss mechanism, which simulates the turning of knobs on a guitar effects pedal, for enhanced accuracy.

This paper is aimed at the usage of an augmented reality assisted system set up on the smart-glasses for training activities. Literature review leads us to a comparison among related technologies, yielding that Mask Regions with Convolutional Neural Network (R-CNN) oriented approach fits the study needs. The proposed method including (1) pointing gesture capture, (2) finger-pointing analysis, and (3) virtual tool positioning and rotation angle are developed. Results show that the recognition of object detection is 95.5%, the Kappa value of recognition of gesture detection is 0.93, and the average time for detecting pointing gesture is 0.26 seconds. Furthermore, even under different lighting, such as indoor and outdoor, the pointing analysis accuracy is up to 79%. The error between the analysis angle and the actual angle is only 1.32 degrees. The results proved that the system is well suited to present the effect of augmented reality, making it applicable for real world usage.

This study proposed a novel system for recognising emotional content in music, and the proposed system is based on particle swarm optimisation (PSO)‐based fuzzy hyper‐rectangular composite neural networks (PFHRCNNs), which integrates three computational intelligence tools, i.e. hyper‐rectangular composite neural networks (HRCNNs), fuzzy systems, and PSO. PFHRCNN is flexible to the complex data due to the fuzzy membership estimation, and an optimisation of the parameters is provided by PSO. First, raw features are extracted from each music clips. After feature extraction, a HRCNN is separately constructed for each class. Each trained HRCNN will result in a set of crisp rules. A problem associated with these generated crisp rules is that some of them are ineffective; therefore, a crisp rule is transformed into a fuzzy rule incorporated with a confidence factor. Next, PSO is adopted to simultaneously trim the rules, search a set of good confidence factors, and fine‐tune the locations of the selected hyper‐rectangles to increase their effectiveness. Finally, a PFHRCNN consisted of a set of fuzzy rules can be generated to recognise the emotion state of music. The experimental result shows that the proposed system has a good performance.

Presents a listing of the Symposium organizing committee.

Summary form only given. At the rehabilitation departments of hospitals in Taiwan, it is not unusual that several patients are assigned to one exercise physiologist or physical therapist during the rehabilitation treatment at a clinical room. One-by-one clinical service is not possible; therefore, an exercise physiologist or a physical therapist cannot know how well his or her patients are practicing the assigned exercises. In addition, patients themselves usually do not have the enough knowledge about whether they are practicing correct therapeutic exercises, leading to ineffective rehabilitation and even produce adverse compensation. Therefore, to alleviate the burdens of exercise physiologists or physical therapists, a therapeutic exercise monitoring system which can automatically measure how well a patient is practicing the suggested exercises deserves to be developed. A good therapeutic exercise monitoring system should be able to provide a measurement report about how well a patient is practicing one suggested exercise; issue a warning alarm signal when an incorrect exercise is detected, etc. For this to happen, it requires an effective motion trajectory recognition algorithm which is able to evaluate whether the therapeutic exercise performed by a patient matches its corresponding assigned exercise. The automatic motion trajectory recognition turns out to be very challenging because motion trajectories are spatial-temporal patterns. There are several surveys on the topics about the domain of human motion analysis. Basically, there are three kinds of approaches (e.g., DTW, HMM, recurrent neural networks, etc.) to spatial-temporal pattern recognition. Each approach has its advantages and limitations. In this talk, a SOM-based motion trajectory recognition algorithm will be introduced. The algorithm starts with the generation of basic posture unit map. The sequence of a motion trajectory is transformed into a 2-D trajectory map based on the basic posture unit map. Then the problem of recognizing motion trajectories is transformed to the problem of recognizing 2-D trajectory maps. Finally, an unknown motion trajectory is classified to be the motion trajectory with the maximum similarity in the motion templates via a template matching technique. The performance of the proposed the SOM-based motion trajectory recognition algorithm is tested on a database consisting of 12 different types of therapeutic exercises. The average 98.8% correct rate was achieved.

Dizziness caused by peripheral vestibular nerve disease might lead to the imbalance of center of gravity, which in turn might affect the quality of daily life. Vestibular function rehabilitation exercise is thought to be able to improve the symptom of dizziness effectively. However, since rehabilitation process is boring, lacks of challenges and adaptive training contents and lacks of assessment method based on exercise analysis, the effectiveness of rehabilitation exercise is thus reduced. In this research, Cawthorne-Cooksey vestibular rehabilitation training exercise in clinical use was modified as an innovative interactive rehabilitation training system in which stereo imagimg, interactive technology and game technology were applied. In this research, clinical test (17 subjects, 6 training sessions) has been successfully carried out, during the experimental process, the task performance data, the time-history of the deviation distribution of center of gravity and subjective feeling questionnaire survey data of the patient is completely measured and recorded. The experimental result has verified the functionalities of this system, in the mean time, through the task performance evaluation and the analysis of level of balance, the medical effect of this system in assisting the patient to perform dizziness rehabilitation training has been proved, and the clinical observation also shows that the patient has higher willingness and motive to use such new system to perform rehabilitation.

In this paper, we introduce a hand-gesture-based control interface for navigating a car-robot. A 3-axis accelerometer is adopted to record a user's hand trajectories. The trajectory data is transmitted wirelessly via an RF module to a computer. The received trajectories are then classified to one of six control commands for navigating a car-robot. The classifier adopts the dynamic time warping (DTW) algorithm to classify hand trajectories. Simulation results show that the classifier could achieve 92.2% correct rate.

Over the past few years there has been an increased interest in developing systems to help drivers drive safely. Although driver fatigue is an important factor in a large number of traffic accidents, a certain percentage of accidents can be due to the inattention factor. In this paper, a low-cost driver inattention monitoring system is proposed. This system can not only detect driver fatigue but also detect whether the driver is looking at other directions for an extended period of time. The system alerts the driver when one of the above two driving conditions happens.

This paper presents a new approach to an on-line learning neuro-fuzzy system based on artificial immune systems. Via the proposed on-line learning algorithm, a neuro-fuzzy system can be incrementally constructed. Several data sets were tested to illustrate the effectiveness of the proposed on-line learning neuro-fuzzy system.

A new approach to optimization problems based on the self-organizing feature maps is proposed. We name the new optimization algorithm the SOM-based optimization (SOMO) algorithm. Through the self-organizing process, good solutions to an optimization problem can be simultaneously explored and exploited. An additional advantage of the algorithm is that the outputs of the neural network allow us to transform a multi-dimensional fitness landscape into a three-dimensional projected fitness landscape. Several simulations are used to illustrate the effectiveness of the proposed optimization algorithm.

We propose a modified version of the K-means algorithm to cluster data. The proposed algorithm adopts a novel nonmetric distance measure based on the idea of "point symmetry". This kind of "point symmetry distance" can be applied in data clustering and human face detection. Several data sets are used to illustrate its effectiveness.

In this paper a new model of self-organizing neural networks is proposed. An algorithm called "double self-organizing feature map" (DSOM) algorithm is developed to train the novel model. By the DSOM algorithm the network will adaptively adjust its network structure during the learning phase so as to make neurons responding to similar stimulus have similar weight vectors and spatially move nearer to each other at the same time. The final network structure allows us to visualize high-dimensional data as a two dimensional scatter plot. The resulting representations allow a straightforward analysis of the inherent structure of clusters within the input data. One high-dimensional data set is used to test the effectiveness of the proposed neural networks.

It is often reported in the technique literature that the success of the self-organizing feature map (SOM) formation is critically dependent on the initial weights and the selection of main parameters of the algorithm, namely, the learning-rate parameter and the neighborhood set. In this paper, we propose a healing mechanism to repair feature maps that are not well topologically ordered. The healed map is then further fine-tuned by the SOM algorithm so as to improve the accuracy of the map. Two data sets are tested to illustrate the performance of the proposed method.

This paper presents a neuro-fuzzy system by using the Kohonen's self-organizing feature map algorithm, not only for its vector quantization feature, but also for its topological property. This property prevents the proposed neuro-fuzzy system from suffering from a drawback like any of the conventional clustering-algorithm-based fuzzy systems, i.e. the optimal number of clusters or some kind of similarity threshold must be predetermined. Associated with the self-organizing feature-map-based fuzzy system is a hybrid learning algorithm, which is for initial parameters setting and fine-tuning the parameters of the system.

We present an efficient approach to forming feature maps. The method involves three stages. In the first stage, we use the K-means algorithm to select N2 (i.e., the size of the feature map to be formed) cluster centers from a data set. Then a heuristic assignment strategy is employed to organize the N2 selected data points into an N x N neural array so as to form an initial feature map. If the initial map is not good enough, then it will be fine-tuned by the traditional Kohonen self-organizing feature map (SOM) algorithm under a fast cooling regime in the third stage. By our three-stage method, a topologically ordered feature map would be formed very quickly instead of requiring a huge amount of iterations to fine-tune the weights toward the density distribution of the data points, which usually happened in the conventional SOM algorithm. Three data sets are utilized to illustrate the proposed method.

Gesture based applications widely range from replacing the traditional mouse as a position device to virtual reality and communication with the deaf. The article presents a fuzzy rule based approach to spatio-temporal hand gesture recognition. This approach employs a powerful method based on hyperrectangutar composite neural networks (HRCNNs) for selecting templates. Templates for each hand shape are represented in the form of crisp IF-THEN rules that are extracted from the values of synaptic weights of the corresponding trained HRCNNs. Each crisp IF-THEN rule is then fuzzified by employing a special membership function in order to represent the degree to which a pattern is similar to the corresponding antecedent part. When an unknown gesture is to be classified, each sample of the unknown gesture is tested by each fuzzy rule. The accumulated similarity associated with all samples of the input is computed for each hand gesture in the vocabulary, and the unknown gesture is classified as the gesture yielding the highest accumulative similarity. Based on the method we can implement a small-sized dynamic hand gesture recognition system. Two databases which consisted of 90 spatio-temporal hand gestures are utilized for verifying its performance. An encouraging experimental result confirms the effectiveness of the proposed method.

In this paper we present an associative-memory-based face detection system. First, the symmetry of human faces is used to quickly locate all the candidates of human faces with all possible sizes and locations. Then two associative memories are used to decide whether or not a human face exists at the locations. Some experimental results are given.

There are many different approaches to recognition of spatio-temporal patterns. Each has its own merits and disadvantages. In this paper we present a neural-network-based approach to spatio-temporal pattern recognition. The effectiveness of this method is evaluated by recognizing 3D arm movements involved in Taiwanese sign language.

We propose a new facial image morphing algorithm based on the Kohonen self-organizing feature map (SOM) algorithm to generate a smooth 2D transformation that reflects anchor point correspondences. Using only a 2D face image and a small number of anchor points, we show that the proposed morphing algorithm provides a powerful mechanism for processing facial expressions.

It is often reported in the technique literature that the success of the self-organizing feature map formation is critically dependent on the initial weights and the selection of main parameters of the algorithm, namely, the learning-rate parameter and the neighborhood function. In this paper, we propose an efficient initialization scheme to construct an initial map. We then use the self-organizing feature map algorithm to make small subsequent adjustments so as to improve the accuracy of the initial map. Two data sets are tested to illustrate the performance of the proposed method.

We present a general approach to deriving a new type of neural network-based fuzzy model for a complex system from numerical and/or linguistic information. To efficiently identify the structure and the parameters of the new fuzzy model, we first partition the output space instead of the input space. As a result, the input space itself induces corresponding partitions within each of which inputs would have similar outputs. Then we use a set of hyperrectangles to fit the partitions of the input space. Consequently, the premise of an implication in the new type of fuzzy rule is represented by a hyperrectangle and the consequence is represented by a fuzzy singleton. A novel two-layer fuzzy hyperrectangular composite neural network (FHRCNN) can be shown to be computationally equivalent to such a special fuzzy model. The process of presenting input data to each hidden node in a FHRCNN is equivalent to firing a fuzzy rule. An efficient learning algorithm was developed to adjust the weights of an FHRCNN. Finally, we apply FHRCNNs to provide real-time transient stability prediction for use with high-speed control in power systems. From simulation tests on the IEEE 39-bus system, it reveals that the proposed novel FHRCNN can yield a much better performance than that of conventional multilayer perceptrons (MLP's) in terms of computational burden and classification rate.

This paper presents a novel class of neural networks which can be trained in an unsupervised manner to detect a mixture of hyperellipsoidal shells and/or segment of hyperellipsoidal shells. This approach is computationally and implementationally simpler than other clustering algorithms that have been suggested for this purpose. Experimental results are given to show the effectiveness of the proposed method.