Current Issue : April - June Volume : 2020 Issue Number : 2 Articles : 5 Articles
Working memory is an important function for human cognition since several day-to-day\nactivities are related to it, such as remembering a direction or developing a mental calculation.\nUnfortunately, working memory deficiencies affect performance in work or education related\nactivities, mainly due to lack of concentration, and, with the goal to improve this, many software\napplications have been developed. However, sometimes the user ends up bored with these games\nand drops out easily. To cope with this, our work explores the use of intelligent robotics and dynamic\ndifficulty adjustment mechanisms to develop a novel working memory training system. The proposed\nsystem, based on the Nao robotic platform, is composed of three main components: First, the N-back\ntask allows stimulating the working memory by remembering visual sequences. Second, a BDI model\nimplements an intelligent agent for decision-making during the progress of the game. Third, a fuzzy\ncontroller, as a dynamic difficulty adjustment system, generates customized levels according to the\nuser. The experimental results of our system, when compared to a computer-based implementation of\nthe N-back game, show a significant improvement on the performance of the user in the game, which\nmight relate to an improvement in their working memory. Additionally, by providing a friendly and\ninteractive interface, the participants have reported a more immersive and better game experience\nwhen using the robotic-based system....
Radio frequency identification (RFID) has shown its potential in humanâ??machine\ninteraction thanks to its inherent function of identification and relevant physical information of\nsignals, but complex data processing and undesirable input accuracy restrict its application and\npromotion in practical use. This paper proposes a novel finger-controlled passive RFID tag design\nfor humanâ??machine interaction. The tag antenna is based on a dipole antenna with a separated Tmatch\nstructure, which is able to adjust the state of the tag by the press of a finger. The state of the\nproposed tag can be recognized directly by the code received by the RFID reader, and no complex\ndata processing is needed. Since the code is hardly affected by surroundings, the proposed tag is\nsuitable to be used as a wireless switch or control button in multiple scenarios. Moreover, arrays of\nthe proposed tag with rational tag arrangements could contribute to a series of manual control\ndevices, such as a wireless keyboard, a remote controller, and a wireless gamepad, without batteries.\nA 3 Ã? 4 array of the finger-controlled tag is presented to constitute a simple passive RFID keyboard\nas an example of the applications of the proposed tag array and it refers to the arrangement of a\nkeypad and can achieve precise, convenient, quick, and practical commands and text input into\nmachines by pressing the tags with fingers. Simulations and measurements of the proposed tag and\ntag array have been carried out to validate their performances in humanâ??machine interaction....
In order to reduce the computational consumption of the training and the testing phases of\nvideo face recognition methods based on a global statistical method and a deep learning network,\na novel video face verification algorithm based on a three-patch local binary pattern (TPLBP) and\nthe 3D Siamese convolutional neural network is proposed in this paper. The proposed method\ntakes the TPLBP texture feature which has excellent performance in face analysis as the input of the\nnetwork. In order to extract the inter-frame information of the video, the texture feature maps of\nthe multi-frames are stacked, and then a shallow Siamese 3D convolutional neural network is used\nto realize dimension reduction. The similarity of high-level features of the video pair is solved by\nthe shallow Siamese 3D convolutional neural network, and then mapped to the interval of 0 to 1 by\nlinear transformation. The classification result can be obtained with the threshold of 0.5. Through an\nexperiment on the YouTube Face database, the proposed algorithm got higher accuracy with less\ncomputational consumption than baseline methods and deep learning methods....
As an important part of emotion research, facial expression recognition is a necessary\nrequirement in humanâ??machine interface. Generally, a face expression recognition system includes\nface detection, feature extraction, and feature classification. Although great success has been made\nby the traditional machine learning methods, most of them have complex computational problems\nand lack the ability to extract comprehensive and abstract features. Deep learning-based methods can\nrealize a higher recognition rate for facial expressions, but a large number of training samples and\ntuning parameters are needed, and the hardware requirement is very high. For the above problems,\nthis paper proposes a method combining features that extracted by the convolutional neural network\n(CNN) with the C4.5 classifier to recognize facial expressions, which not only can address the\nincompleteness of handcrafted features but also can avoid the high hardware configuration in the\ndeep learning model. Considering some problems of overfitting and weak generalization ability\nof the single classifier, random forest is applied in this paper. Meanwhile, this paper makes some\nimprovements for C4.5 classifier and the traditional random forest in the process of experiments.\nA large number of experiments have proved the effectiveness and feasibility of the proposed method....
The development of advanced technologies for wireless data collection and the analysis\nof quantitative data, with application to a humanâ??machine interface (HMI), is of growing interest.\nIn particular, various wearable devices related to HMIs are being developed. These devices require a\ncustomization process that considers the physical characteristics of each individual, such as mounting\npositions of electrodes, muscle masses, and so forth. Here, the authors report device and calculation\nconcepts for flexible platforms that can measure electrical signals changed through electromyography\n(EMG). This soft, flexible, and lightweight EMG sensor can be attached to curved surfaces such as\nthe forearm, biceps, back, legs, etc., and optimized biosignals can be obtained continuously through\npost-processing. In addition to the measurement of EMG signals, the application of the HMI has\nstable performance and high accuracy of more than 95%, as confirmed by 50 trials per case. The result\nof this study shows the possibility of application to various fields such as entertainment, the military,\nrobotics, and healthcare in the future....
Loading....