This paper presents a smartphone app that performs real-time voice activity detection based on convolutional neural network. Real-time implementation issues are discussed showing how the slow inference time associated with convolutional neural networks is addressed. The developed smartphone app is meant to act as a switch for noise reduction in the signal processing pipelines of hearing devices, enabling noise estimation or classification to be conducted in noise-only parts of noisy speech signals. The developed smartphone app is compared with a previously developed voice activity detection app as well as with two highly cited voice activity detection algorithms. The experimental results indicate that the developed app using convolutional neural network outperforms the previously developed smartphone app.
This paper presents the real-time implementation and field testing of an app running on smartphones for classifying noise signals involving subband features and a random forest classifier. This app is compared to a previously developed app utilizing mel-frequency cepstral coefficients features and a Gaussian mixture model classifier. The real-time implementation has been carried out on both the Android and iOS smartphones. The field testing results indicate the superiority of this newly developed app over the previously developed app in terms of classification rates.
19 March 2024 in IEEE Transactions on Wireless Communications
Due to its ubiquitous and contact-free nature, the use of WiFi infrastructure for performing sensing tasks has tremendous potential. However, the channel state information (CSI) measured by a WiFi receiver suffers from errors in both its gain and phase, which can significantly hinder sensing tasks. By analyzing these errors from different WiFi receivers, a mathematical model for these gain and phase errors is developed in this work. Based on these models, several theoretically justified preprocessing algorithms for correcting such errors at a receiver and, thus, obtaining clean CSI are presented. Simulation results show that at typical system parameters, the developed algorithms for cleaning CSI can reduce noise by 40% and 200%, respectively, compared to baseline methods for gain correction and phase correction, without significantly impacting computational cost. The superiority of the proposed methods is also validated in a real-world test bed for respiration rate monitoring (an example sensing task), where they improve the estimation signal-to-noise ratio by 20% compared to baseline methods.
Deep learning solutions are being increasingly used in mobile applications. Although there are many open-source software tools for the development of deep learning solutions, there are no guidelines in one place in a unified manner for using these tools toward real-time deployment of these solutions on smartphones. From the variety of available deep learning tools, the most suited ones are used in this paper to enable real-time deployment of deep learning inference networks on smartphones. A uniform flow of implementation is devised for both Android and iOS smartphones. The advantage of using multi-threading to achieve or improve real-time throughputs is also showcased. A benchmarking framework consisting of accuracy, CPU/GPU consumption, and real-time throughput is considered for validation purposes. The developed deployment approach allows deep learning models to be turned into real-time smartphone apps with ease based on publicly available deep learning and smartphone software tools. This approach is applied to six popular or representative convolutional neural network models, and the validation results based on the benchmarking metrics are reported.
IEEE 802.11ax newly introduced Target Wake Time (TWT) function which enables joint reduction of device power consumption and wireless medium congestion through negotiated TWT wake interval and duration. We develop a novel solution which can dynamically configure station (STA) wake interval and duration based on run time recognition of the Quality of Service (QoS) requirement including the required latency and throughput. Specifically, a Machine Learning (ML) based Network Service Detection (NSD) module is developed which can detect the current network service in real time and update the STA wake interval according to the corresponding service latency requirement. Moreover, a novel data time estimation model is developed which can estimate the required throughput and wake duration with the observed throughput, linkspeed, and contention level. In addition, a state-machine based method is developed to detect three different traffic types (random, stable, bursty) and their throughput variation pattern for further optimization of the STA wake duration. Our solution is implemented and extensively tested on commercial smart mobile platform under various network conditions and use cases. The results show that our ML based NSD could reach 99.2% and 96.5% accuracy respectively for coarse-grain and fine-grain network service classification, which ensures our TWT solution can accurately recognize and satisfy the QoS requirements. More importantly, while maintaining the QoS and user experience, our solution can substantially reduce the Wi-Fi duty cycle to 29.6% on average, which leads to lower device power consumption and network contention level.
This papers presents a low-latency smartphone app to achieve real-time noise reduction of speech signals in noisy sound environments. This app overcomes the two shortcomings of high latency and musical noise artifact that are associated with the previously developed apps for the same purpose. The results of both objective and subjective evaluations are reported which show the effectiveness of this app towards reducing noise for speech signals in noisy environments, in particular in babble background noise.