Deep-learning-based speech enhancement techniques have been recently grown in interest, since their impressive performance can potentially benefit a wide variety of digital voice communication systems. However, such performance has been evaluated mostly in offline audio processing scenarios (i.e. feeding the model, in one go, a complete audio recording, which may extend several seconds). It is of great interest to evaluate and characterize the current state-of-the-art in applications that process audio online (i.e. feeding the model a sequence of segments of audio data, concatenating the results at the output end). Although evaluations and comparisons between speech enhancement techniques have been carried out before, as far as the author knows, the work presented here is the first that evaluates the performance of such techniques in relation to their online applicability. Meaning, this work measures how the output signal-to-interference ratio (as a separation metric), the response time and memory usage (as online metrics) are impacted by the input length (the size of audio segments), in addition to the amount of noise, amount and number of interferences, and amount of reverberation. Three popular models were evaluated, given their availability on public repositories and online viability: MetricGAN+, Spectral Feature Mapping with Mimic Loss, and Demucs-Denoiser. The characterization was carried out using a systematic evaluation protocol based on the Speechbrain framework. Several intuitions are presented and discussed, and some recommendations for future work are proposed.

Online speech enhancement is of great interest, because of its diverse areas of application. The Demucs architecture has been recently shown to achieve a high level of performance, even with small audio segments, while requiring a relatively small amount of computational resources. However, an issue that it bears, as well as most of state-of-the-art speech enhancement techniques, is that of target speech selection: it is assumed that only one speech source is present. However, in real-life scenarios, more than one speech source may be present, in it is uncertain which speech source within the fed mixture should be enhanced. Thus, it is of interest to complement speech enhancement techniques (such as Demucs-based) with a target selection scheme, so as to ensure that only the source of interest is enhanced. In this work, two target selection strategies are explored: 1) an embedding-based strategy, using a codified sample of the target speech, and 2) a location-based strategy, using a beamforming-based pre-filter to select the target that is in front of a 2-microphone array. It is shown that while both strategies improve the performance of the Demucs-based technique when one or more speech interferences are present, they both have their pros and cons. Specifically, while the beamforming-based strategy achieves overall a better performance compared to the embedding-based strategy, it is sensitive against the location variation of the target speech source which the embedding-based strategy does not suffer from.

Acoustic energy mapping provides the functionality to obtain characteristics of acoustic sources, such as: presence, localization, type and trajectory of sound sources. Several beamforming-based techniques can be used for this purpose, however, they rely on the difference of arrival times of the signal at each capture node (or microphone), so it is of major importance to have synchronized multi-channel recordings. A Wireless Acoustic Sensor Network (WASN) can be very practical to install when used for mapping the acoustic energy of a given acoustic environment. However, they are known for having low synchronization between the recordings from each node. The objective of this paper is to characterize the impact of current popular synchronization methodologies as part of the WASN to capture reliable data to be used for acoustic energy mapping. The two evaluated synchronization protocols are: Network Time Protocol (NTP) y Precision Time Protocol (PTP). Additionally, three different audio capture methodologies were proposed for the WASN to capture the acoustic signal: two of them, recording the data locally and one sending the data through a local wireless network. As a real-life evaluation scenario, a WASN was built using nodes conformed by a Raspberry Pi 4B+ with a single MEMS microphone. Experimental results demonstrate that the most reliable methodology is using the PTP synchronization protocol and audio recording locally.