This work is a preliminary study towards developing an alternative communication channel for conveying shape information to aid in recognition of items when tactile perception is hindered. Tactile data, acquired during object exploration by sensor fitted robot arm, are processed to recognize four basic geometric shapes. Patterns representing each shape, classified from tactile data, are generated using micro-controller-driven vibration motors which vibrotactually stimulate users to convey the particular shape information. These motors are attached on the subject's arm and their psychological (verbal) responses are recorded to assess the competence of the system to convey shape information to the user in form of vibrotactile stimulations. Object shapes are classified from tactile data with an average accuracy of 95.21 %. Three successive sessions of shape recognition from vibrotactile pattern depicted learning of the stimulus from subjects' psychological response which increased from 75 to 95 %. This observation substantiates the learning of vibrotactile stimulation in user over the sessions which in turn increase the system efficacy. The tactile sensing module and vibrotactile pattern generating module are integrated to complete the system whose operation is analysed in real-time. Thus, the work demonstrates a successful implementation of the complete schema of artificial tactile sensing system for object-shape recognition through vibrotactile stimulations.
This paper proposes a new classification algorithm which aims at predicting different states from an incoming non-stationary signal. To overcome the failure of standard classifiers at generalizing the patterns for such signals, we have proposed an Interval Type-2 Fuzzy based Adaptive neural fuzzy Inference System (ANFIS). Through the introduction IT2F system, we have aimed at improving the uncertainty management of the fuzzy inference system. Besides that using DE in forward and backward pass and improving the forward pass function we have improved the parameter update on wide range of nodal functions without any quadratic approximation in forward pass. The proposed algorithm is tested on a standard electroencephalography (EEG) dataset and it is noted that the proposed algorithm performs better than other standard classifiers including the classical ANFIS algorithm.
Superposition of noise with the bio-potential signals causes plethoric information loss and misinterpretation in human-computer interference systems. Here, we have proposed a novel method to design a digital signal filter which is capable of filtering four major bio-potential signals viz. Electroencephalography (EEG), Electrooculography (EOG), Electrocardiography (ECG) and Electromyography (EMG). Different sampling frequencies for different bio-potential signals are manually selected through two select lines added as external inputs to the microcontroller based system. The filter has two sections in cascade, a conventional sixth order digital filter followed by an adaptive interference canceller (AIC). The AIC has a remarkable property of eliminating most of the interfering power line noise adaptively without consuming much time or space complexity of the embedded processor used. A variant of Least Mean Square (LMS) filtering algorithm called Amplitude-Phase Adaptive LMS (APALMS) is implemented here. Convergence behaviors of the adaptive parameters are simulated and finally verified on real time biopotential signals extracted from different subjects using low cost embedded processor. In this paper, we have found that our proposed filter transition frequencies are small with least variation in adaptation time.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.