Speech separation is the task of separating target speech from background interference. Traditionally, speech separation is studied as a signal processing problem. A more recent approach formulates speech separation as a supervised learning problem, where the discriminative patterns of speech, speakers, and background noise are learned from training data. Over the past decade, many supervised separation algorithms have been put forward. In particular, the recent introduction of deep learning to supervised speech separation has dramatically accelerated progress and boosted separation performance. This paper provides a comprehensive overview of the research on deep learning based supervised speech separation in the last several years. We first introduce the background of speech separation and the formulation of supervised separation. Then, we discuss three main components of supervised separation: learning machines, training targets, and acoustic features. Much of the overview is on separation algorithms where we review monaural methods, including speech enhancement (speech-nonspeech separation), speaker separation (multitalker separation), and speech dereverberation, as well as multimicrophone techniques. The important issue of generalization, unique to supervised learning, is discussed. This overview provides a historical perspective on how advances are made. In addition, we discuss a number of conceptual issues, including what constitutes the target source.

Supervised Speech Separation Based on Deep Learning: An Overview

Cellular systems evolved from a dedicated mobile communication system to an almost omnipresent system with unlimited coverage anywhere and anytime for any device. The growing ubiquity of the network stirred expectations to determine the location of the mobile devices themselves. Since the beginning of standardization, each cellular mobile radio generation has been designed for communication services, and satellite navigation systems, such as Global Positioning System (GPS), have provided precise localization as an add-on service to the mobile terminal. Self-contained localization services relying on the mobile network elements have offered only rough position estimates. Moreover, satellite-based technologies suffer a severe degradation of their localization performance in indoors and urban areas. Therefore, only in subsequent cellular standard releases, more accurate cellular-based location methods have been considered to accommodate more challenging localization services. This survey provides an overview of the evolution of the various localization methods that were standardized from the first to the fourth generation of cellular mobile radio, and looks over what can be expected with the new radio and network aspects for the upcoming generation of fifth generation.

Survey of Cellular Mobile Radio Localization Methods: From 1G to 5G

Speech separation is the task of separating target speech from background interference. Traditionally, speech separation is studied as a signal processing problem. A more recent approach formulates speech separation as a supervised learning problem, where the discriminative patterns of speech, speakers, and background noise are learned from training data. Over the past decade, many supervised separation algorithms have been put forward. In particular, the recent introduction of deep learning to supervised speech separation has dramatically accelerated progress and boosted separation performance. This article provides a comprehensive overview of the research on deep learning based supervised speech separation in the last several years. We first introduce the background of speech separation and the formulation of supervised separation. Then we discuss three main components of supervised separation: learning machines, training targets, and acoustic features. Much of the overview is on separation algorithms where we review monaural methods, including speech enhancement (speech-nonspeech separation), speaker separation (multi-talker separation), and speech dereverberation, as well as multi-microphone techniques. The important issue of generalization, unique to supervised learning, is discussed. This overview provides a historical perspective on how advances are made. In addition, we discuss a number of conceptual issues, including what constitutes the target source.

Supervised Speech Separation Based on Deep Learning: An Overview.

Fundamentals of statistical signal processing: Estimation theory: by Steven M. KAY; Prentice Hall signal processing series; Prentice Hall; Englewood Cliffs, NJ, USA; 1993; xii + 595 pp.; $65; ISBN: 0-13-345711-7

Phase is important for perceptual quality of speech. However, it seems intractable to directly estimate phase spectra through supervised learning due to their lack of spectrotemporal structure in it. Complex spectral mapping aims to estimate the real and imaginary spectrograms of clean speech from those of noisy speech, which simultaneously enhances magnitude and phase responses of speech. Inspired by multi-task learning, we propose a gated convolutional recurrent network (GCRN) for complex spectral mapping, which amounts to a causal system for monaural speech enhancement. Our experimental results suggest that the proposed GCRN substantially outperforms an existing convolutional neural network (CNN) for complex spectral mapping in terms of both objective speech intelligibility and quality. Moreover, the proposed approach yields significantly higher STOI and PESQ than magnitude spectral mapping and complex ratio masking. We also find that complex spectral mapping with the proposed GCRN provides an effective phase estimate.

Learning Complex Spectral Mapping With Gated Convolutional Recurrent Networks for Monaural Speech Enhancement

Robust indoor positioning and location awareness at a sub-meter accuracy typically require highly accurate radio channel measurements to extract precise time-of-flight measurements. Emerging UWB transponders like the DecaWave DW1000 chip offer to estimate channel impulse responses with reasonably high bandwidth and excellent clock stability, yielding a ranging precision below 10 cm. The competitive pricing of these chips allows scientists and engineers for the first time to exploit the benefits of UWB for indoor positioning without the need for a massive investment into experimental equipment. This work investigates the performance of the DW1000 chip concerning position related information that can be extracted from its channel impulse response measurements. We evaluate the signal-to-interference-plus-noise ratio of the line-of-sight and reflected multipath components which is a key parameter determining the Cramer-Rao lower bound on the ranging error variance. We propose a novel and highly efficient positioning algorithm, which requires information from a single anchor only. Results demonstrate reliable and robust positioning at an accuracy below 0.5 m.

Using DecaWave UWB transceivers for high-accuracy multipath-assisted indoor positioning

Highly accurate and reliable indoor positioning — at accuracy levels in the 10 cm range — will enable a large a number of innovative location-based applications because such accuracy levels essentially allow for a useful real-time interaction of humans and cyber-physical systems. Activity recognition, navigation at "shelf" level, geofencing, process monitoring and process control are among the envisioned services that will yield numerous applications in various domains. This paper reviews the difficulties faced by indoor positioning systems, motivating the requirement for a large signal bandwidth and how a lack of bandwidth can be compensated by multi-antenna systems. The potential capabilities of upcoming generations of wireless systems will increasingly make high-accuracy positioning available in near future.

High-accuracy positioning for indoor applications: RFID, UWB, 5G, and beyond

In conventional single-channel speech enhancement, typically the noisy spectral amplitude is modified while the noisy phase is used to reconstruct the enhanced signal. Several recent attempts have shown the effectiveness of utilizing an improved spectral phase for phase-aware speech enhancement and consequently its positive impact on the perceived speech quality. In this paper, we present a harmonic phase estimation method relying on fundamental frequency and signal-to-noise ratio (SNR) information estimated from noisy speech. The proposed method relies on SNR-based time-frequency smoothing of the unwrapped phase obtained from the decomposition of the noisy phase. To incorporate the uncertainty in the estimated phase due to unreliable voicing decision and SNR estimate, we propose a binary hypothesis test assuming speech-present and speech-absent classes representing high and low SNRs. The effectiveness of the proposed phase estimation method is evaluated for both phase-only enhancement of noisy speech and in combination with an amplitude-only enhancement scheme. We show that by enhancing the noisy phase both perceived speech quality as well as speech intelligibility are improved as predicted by the instrumental metrics and justified by subjective listening tests.

Harmonic phase estimation in single-channel speech enhancement using phase decomposition and SNR information

Setting up indoor localization systems is often excessively time-consuming and labor-intensive, because of the high amount of anchors to be carefully deployed or the burdensome collection of fingerprints. In this paper, we present SALMA, a novel low-cost UWB-based indoor localization system that makes use of only one anchor and that does neither require prior calibration nor training. By using only a crude floor plan and by exploiting multipath reflections, SALMA can accurately determine the position of a mobile tag using a single anchor, hence minimizing the infrastructure costs, as well as the setup time. We implement SALMA on off-the-shelf UWB devices based on the Decawave DW1000 transceiver and show that, by making use of multiple directional antennas, SALMA can also resolve ambiguities due to overlapping multipath components. An experimental evaluation in an office environment with clear line-of-sight has shown that 90% of the position estimates obtained using SALMA exhibit less than 20 cm error, with a median below 8 cm. We further study the performance of SALMA in the presence of obstructed line-of-sight conditions, moving objects and furniture, as well as in highly dynamic environments with several people moving around, showing that the system can sustain decimeter-level accuracy with a worst-case average error below 34 cm.

SALMA: UWB-based Single-Anchor Localization System using Multipath Assistance

5G mmWave communication is useful for positioning due to the geometric connection between the propagation channel and the propagation environment. Channel estimation methods can exploit the resulting sparsity to estimate parameters (delay and angles) of each propagation path, which in turn can be exploited for positioning and mapping. When paths exhibit significant spread in either angle or delay, these methods break down or lead to significant biases. We present a novel tensor-based method for channel estimation that allows estimation of mmWave channel parameters in a non-parametric form. The method is able to accurately estimate the channel, even in the absence of a specular component. This in turn enables positioning and mapping using only diffuse multipath. Simulation results are provided to demonstrate the efficacy of the proposed approach.

/pdf/5g-positioning-and-mapping-with-diffuse-multipath-32j0v4c4of.pdf

Josef Kulmer

Papers

Using DecaWave UWB transceivers for high-accuracy multipath-assisted indoor positioning

High-accuracy positioning for indoor applications: RFID, UWB, 5G, and beyond

Harmonic phase estimation in single-channel speech enhancement using phase decomposition and SNR information

SALMA: UWB-based Single-Anchor Localization System using Multipath Assistance

5G Positioning and Mapping With Diffuse Multipath