In many speech applications such as hands-free telephony or voice-controlled home assistants, the distance between the user and the recording microphones can be relatively large. In such a far-field scenario, the recorded microphone signals are typically corrupted by noise and reverberation, which may severely degrade the performance of speech recognition systems and reduce intelligibility and quality of speech in communication applications. In order to limit these effects, speech enhancement algorithms are typically applied. The main objective of this thesis is to develop novel speech enhancement algorithms for noisy and reverberant environments and signal-based measures to evaluate these algorithms, focusing on solutions that are applicable in realistic scenarios. First, we propose a single-channel speech enhancement algorithm for joint noise and reverberation reduction. The proposed algorithm uses a spectral gain to enhance the input signal, where the gain is computed using a ...

In natural listening environments, speech signals are easily distorted by variousacoustic interference, which reduces the speech quality and intelligibility of human listening; meanwhile, it makes difficult for many speech-related applications, such as automatic speech recognition (ASR). Thus, many speech enhancement (SE) algorithms have been developed in the past decades. However, most current SE algorithms are difficult to capture underlying speech information (e.g., phoneme) in the SE process. This causes it to be challenging to know what specific information is lost or interfered with in the SE process, which limits the application of enhanced speech. For instance, some SE algorithms aimed to improve human listening usually damage the ASR system. The objective of this dissertation is to develop SE algorithms that have the potential to capture various underlying speech representations (information) and improve the quality and intelligibility of noisy speech. This ...

The ability to communicate through speech is important for social interaction. We rely on the ability to communicate with each other even in noisy conditions. Ideally, the speech is easy to understand but this is not always the case, if the speech is degraded, e.g., due to background noise, distortion or hearing impairment. One of the most important factors to consider in relation to such degradations is speech intelligibility, which is a measure of how easy or difficult it is to understand the speech. In this thesis, the focus is on the topic of speech intelligibility prediction. The thesis consists of an introduction to the field of speech intelligibility prediction and a collection of scientific papers. The introduction provides a background to the challenges with speech communication in noisy conditions, followed by an introduction to how speech is produced and ...

The increasing use of technological devices and biometric recognition systems in people daily lives has motivated a great deal of research interest in the development of effective and robust systems. However, there are still some challenges to be solved in these systems when Deep Neural Networks (DNNs) are employed. For this reason, this thesis proposes different approaches to address these issues. First of all, we have analyzed the effect of introducing the most widespread DNN architectures to develop systems for face and text-dependent speaker verification tasks. In this analysis, we observed that state-of-the-art DNNs established for many tasks, including face verification, did not perform efficiently for text-dependent speaker verification. Therefore, we have conducted a study to find the cause of this poor performance and we have noted that under certain circumstances this problem is due to the use of a ...

Speech signals acquired, transmitted or stored in non-ideal conditions are often degraded by one or more effects including, for example, additive noise. These degradations alter the signal properties in a manner that deteriorates the intelligibility or quality of the speech signal. In the law enforcement context such degradations are commonplace due to the limitations in the audio collection methodology, which is often required to be covert. In severe degradation conditions, the acquired signal may become unintelligible, losing its value in an investigation and in less severe conditions, a loss in signal quality may be encountered, which can lead to higher transcription time and cost. This thesis proposes a non-intrusive speech assessment framework from which algorithms for speech quality and intelligibility assessment are derived, to guide the collection and transcription of law enforcement audio. These methods are trained on a large ...

The subject of this PhD thesis is the efficient and robust processing and analysis of the audio recordings that are derived from a call center. The thesis is comprised of two parts. The first part is dedicated to dialogue/non-dialogue detection and to speaker segmentation. The systems that are developed are prerequisite for detecting (i) the audio segments that actually contain a dialogue between the system and the call center customer and (ii) the change points between the system and the customer. This way the volume of the audio recordings that need to be processed is significantly reduced, while the system is automated. To detect the presence of a dialogue several systems are developed. This is the first effort found in the international literature that the audio channel is exclusively exploited. Also, it is the first time that the speaker utterance ...

The recent progress in artificial neural networks (rebranded as “deep learning”) has significantly boosted the state-of-the-art in numerous domains of computer vision offering an opportunity to approach the problems which were hardly solvable with conventional machine learning. Thus, in the frame of this PhD study, we explore how deep learning techniques can help in the analysis of one the most basic and essential semantic traits revealed by a human face, namely, gender and age. In particular, two complementary problem settings are considered: (1) gender/age prediction from given face images, and (2) synthesis and editing of human faces with the required gender/age attributes. Convolutional Neural Network (CNN) has currently become a standard model for image-based object recognition in general, and therefore, is a natural choice for addressing the first of these two problems. However, our preliminary studies have shown that the ...

A large part of human speech communication takes place in noisy environments and is supported by technical devices. For example, a hearing-impaired person might use a hearing aid to take part in a conversation in a busy restaurant. These devices, but also telecommunication in noisy environments or voiced-controlled assistants, make use of speech enhancement and separation algorithms that improve the quality and intelligibility of speech by separating speakers and suppressing background noise as well as other unwanted effects such as reverberation. If the devices are equipped with more than one microphone, which is very common nowadays, then multi-channel speech enhancement approaches can leverage spatial information in addition to single-channel tempo-spectral information to perform the task. Traditionally, linear spatial filters, so-called beamformers, have been employed to suppress the signal components from other than the target direction and thereby enhance the desired ...

This thesis contributes to motion characterization through inertial and physiological signals captured by wearable devices and analyzed using signal processing and deep learning techniques. This research leverages the possibilities of motion analysis for three main applications: to know what physical activity a person is performing (Human Activity Recognition), to identify who is performing that motion (user identification) or know how the movement is being performed (motor anomaly detection). Most previous research has addressed human motion modeling using invasive sensors in contact with the user or intrusive sensors that modify the user’s behavior while performing an action (cameras or microphones). In this sense, wearable devices such as smartphones and smartwatches can collect motion signals from users during their daily lives in a less invasive or intrusive way. Recently, there has been an exponential increase in research focused on inertial-signal processing to ...

Deep learning is an emerging technology that is considered one of the most promising directions for reaching higher levels of artificial intelligence. Among the other achievements, building computers that understand speech represents a crucial leap towards intelligent machines. Despite the great efforts of the past decades, however, a natural and robust human-machine speech interaction still appears to be out of reach, especially when users interact with a distant microphone in noisy and reverberant environments. The latter disturbances severely hamper the intelligibility of a speech signal, making Distant Speech Recognition (DSR) one of the major open challenges in the field. This thesis addresses the latter scenario and proposes some novel techniques, architectures, and algorithms to improve the robustness of distant-talking acoustic models. We first elaborate on methodologies for realistic data contamination, with a particular emphasis on DNN training with simulated data. ...

The explosion of Deep Learning (DL) added a boost to the already rapidly developing field of Computer Vision to such a point that vision-based tasks are now parts of our everyday lives. Applications such as image classification, photo stylization, or face recognition are nowadays pervasive, as evidenced by the advent of modern systems trivially integrated into mobile applications. In this thesis, we investigated and enhanced the visual counting task, which automatically estimates the number of objects in still images or video frames. Recently, due to the growing interest in it, several Convolutional Neural Network (CNN)-based solutions have been suggested by the scientific community. These artificial neural networks, inspired by the organization of the animal visual cortex, provide a way to automatically learn effective representations from raw visual data and can be successfully employed to address typical challenges characterizing this task, ...

The problem of segregating a sound source of interest from an acoustic background has been extensively studied due to applications in hearing prostheses, robust speech/speaker recognition and audio information retrieval. Computational auditory scene analysis (CASA) approaches the segregation problem by utilizing grouping cues involved in the perceptual organization of sound by human listeners. Binaural processing, where input signals resemble those that enter the two ears, is of particular interest in the CASA field. The dominant approach to binaural segregation has been to derive spatially selective filters in order to enhance the signal in a direction of interest. As such, the problems of sound localization and sound segregation are closely tied. While spatial filtering has been widely utilized, substantial performance degradation is incurred in reverberant environments and more fundamentally, segregation cannot be performed without sufficient spatial separation between sources. This dissertation ...

Over the last few years, i-vectors have been the state-of-the-art technique in speaker and language recognition. Recent advances in Deep Learning (DL) technology have improved the quality of i-vectors but the DL techniques in use are computationally expensive and need speaker or/and phonetic labels for the background data, which are not easily accessible in practice. On the other hand, the lack of speaker-labeled background data makes a big performance gap, in speaker recognition, between two well-known cosine and Probabilistic Linear Discriminant Analysis (PLDA) i-vector scoring techniques. It has recently been a challenge how to fill this gap without speaker labels, which are expensive in practice. Although some unsupervised clustering techniques are proposed to estimate the speaker labels, they cannot accurately estimate the labels. This thesis tries to solve the problems above by using the DL technology in different ways, without ...

Much like natural language, music is highly structured, with strong priors on the likelihood of note sequences. In automatic speech recognition (ASR), these priors are called language models, which are used in addition to acoustic models and participate greatly to the success of today's systems. However, in Automatic Music Transcription (AMT), ASR's musical equivalent, Music Language Models (MLMs) are rarely used. AMT can be defined as the process of extracting a symbolic representation from an audio signal, describing which notes were played at what time. In this thesis, we investigate the design of MLMs using recurrent neural networks (RNNs) and their use for AMT. We first look into MLM performance on a polyphonic prediction task. We observe that using musically-relevant timesteps results in desirable MLM behaviour, which is not reflected in usual evaluation metrics. We compare our model against benchmark ...

The production and delivery of audio for television involve many creative and technical challenges. One of them is concerned with the level balance between the foreground speech (also referred to as dialogue) and the background elements, e.g., music, sound effects, and ambient sounds. Background elements are fundamental for the narrative and for creating an engaging atmosphere, but they can mask the dialogue, which the audience wishes to follow in a comfortable way. Very different individual factors of the people in the audience clash with the creative freedom of the content creators. As a result, service providers receive regular complaints about difficulties in understanding the dialogue because of too loud background sounds. While this has been a known issue for at least three decades, works analyzing the problem and up-to-date statics were scarce before the contributions in this work. Enabling the ...