Human Auditory Mechanism Research Articles

Robust voice activity detection using an auditory-inspired masked modulation encoder based convolutional attention network

Deep learning has revolutionized voice activity detection (VAD) by offering promising solutions. However, directly applying traditional features, such as raw waveforms and Mel-frequency cepstral coefficients, to deep neural networks often leads to degraded VAD performance due to noise interference. In contrast, humans possess the remarkable ability to discern speech in complex and noisy environments, which motivated us to draw inspiration from the human auditory system. We propose a robust VAD algorithm called auditory-inspired masked modulation encoder based convolutional attention network (AMME-CANet) that integrates our AMME with CANet. Firstly, we investigate the design of auditory-inspired modulation features as a deep-learning encoder (AME), effectively simulating the process of sound-signal transmission to inner ear hair cells and subsequent modulation filtering by neural cells. Secondly, building upon the observed masking effects in the human auditory system, we enhance our auditory-inspired modulation encoder by incorporating a masking mechanism resulting in the AMME. The AMME amplifies cleaner speech frequencies while suppressing noise components. Thirdly, inspired by the human auditory mechanism and capitalizing on contextual information, we leverage the attention mechanism for VAD. This methodology uses an attention mechanism to assign higher weights to contextual information containing richer and more informative cues. Through extensive experimentation and evaluation, we demonstrated the superior performance of AMME-CANet in enhancing VAD under challenging noise conditions.

An intelligent healthcare system for detection and classification to discriminate vocal fold disorders

The growing population of senior citizens around the world will appear as a big challenge in the future and they will engage a significant portion of the healthcare facilities. Therefore, it is necessary to develop intelligent healthcare systems so that they can be deployed in smart homes and cities for remote diagnosis. To overcome the problem, an intelligent healthcare system is proposed in this study. The proposed intelligent system is based on the human auditory mechanism and capable of detection and classification of various types of the vocal fold disorders. In the proposed system, critical bandwidth phenomena by using the bandpass filters spaced over Bark scale is implemented to simulate the human auditory mechanism. Therefore, the system acts like an expert clinician who can evaluate the voice of a patient by auditory perception. The experimental results show that the proposed system can detect the pathology with an accuracy of 99.72%. Moreover, the classification accuracy for vocal fold polyp, keratosis, vocal fold paralysis, vocal fold nodules, and adductor spasmodic dysphonia is 97.54%, 99.08%, 96.75%, 98.65%, 95.83%, and 95.83%, respectively. In addition, an experiment for paralysis versus all other disorders is also conducted, and an accuracy of 99.13% is achieved. The results show that the proposed system is accurate and reliable in vocal fold disorder assessment and can be deployed successfully for remote diagnosis. Moreover, the performance of the proposed system is better as compared to existing disorder assessment systems.

승용차 디젤 엔진 소음에 대한 음질 평가 기법 연구

The brand sound of vehicle diesel engine is recently one of the important advantage strategies in the automotive company. Because various noise components masked under high frequency level can be audible in quieter driving situation. Many researches have been carried out for subjective and objective assessments on vehicle sounds and noises. In particular, the interior sound quality has been one of research fields that can give high quality feature to vehicle products. Vehicle interior noise above 500 Hz is usually controlled by sound package parts. The materials and geometries of sound package parts directly affect on this high frequency noise. This paper describes the sound quality evaluation method for the vehicle diesel engine noise to establish objective criteria for sound quality assessment. Considering the sensitivity of human hearing to impulsive sounds such as diesel noise, the human auditory mechanism was simulated by introducing temporal masking in the time domain. Furthermore, each of the human auditory organs was simulated by computer codes, providing reasonable analytical explanations of typical human hearing responses to diesel noise. This method finally provides the sound quality index of vehicle diesel engine noise that includes high frequency intermittent offensive sounds caused by impacting excitations of combustion and piston slap.

Relation between binaural phase and amplitude differences in the perception of movement of a fused, lateralized sound image

A series of experiments investigated the interaction of phase and amplitude modulation in the auditory perception of movement of a fused, lateralized sound image. A technique was developed whereby the perception of periodic motion in the coronal plane of the head could be produced either by sinusoidal modulation of the amplitude or phase relation of a pure tone carrier presented to each ear. At a given modulation frequency the observer set the amplitude of either the phase shift or the differential amplitude change in order to just detect motion in the sound image. The threshold amount of each measure was defined as unity and the observer was then presented with various ratios of the two mixed in two ways: (1) Phase lead to one ear with amplitude increase to the same ear or (2) phase lead to one ear with amplitude increase to the other ear. Evidence for super additivity of amplitude and phase modulation was found for condition I while evidence for subadditivity for am and pm was found in condition number 2. Both these results suggest that, although both time and intensity differences are used by the human auditory mechanisms involved in the perception of movement, they have at least some different neural substrates.

Attenuation Characteristics of Recreational Helmets

The noise levels of many off-the-road vehicles (ORVs) are potentially capable of causing damage to the human auditory mechanism. Most drivers of these machines refuse to wear ear protection because they contend that their helmets provide adequate attenuation. This study was designed to evaluate the attenuation of two typical recreational helmets. The results of this investigation indicated that recreational headgear alone is not sufficient protection from high noise levels. Drivers of ORVs should wear some form of ear protection in conjunction with their headgear until the noise levels of these machines are reduced.

An electronic random number generator

THIS ARTICLE DISCUSSES the generation of random numbers for use in auditory psychophysical experiments conducted at the University of Michigan's Electronic Defense Group laboratories. These experiments, which require randomized signal presentation, are designed to test theoretical hypotheses related to the operation of the human auditory mechanism.