This book covers the state-of-the-art in deep neural-network-based methods for noise robustness in distant speech recognition applications. It provides insights and detailed descriptions of some of the new concepts and key technologies in the field, including novel architectures for speech enhancement, microphone arrays, robust features, acoustic model adaptation, training data augmentation, and training criteria. The contributed chapters also include descriptions of real-world applications, benchmark tools and datasets widely used in the field. This book is intended for researchers and practitioners working in the field of speech processing and recognition who are interested in the latest deep learning techniques for noise robustness. It will also be of interest to graduate students in electrical engineering or computer science, who will find it a useful guide to this field of research.
This book constitutes the proceedings of the 22nd International Conference on Text, Speech, and Dialogue, TSD 2019, held in Ljubljana, Slovenia, in September 2019. The 33 full papers presented in this volume were carefully reviewed and selected from 73 submissions. They were organized in topical sections named text and speech. The book also contains one invited talk in full paper length.
This book constitutes the proceedings of the 7th International Conference on Statistical Language and Speech Processing, SLSP 2019, held in Ljubljana, Slovenia, in October 2019. The 25 full papers presented together with one invited paper in this volume were carefully reviewed and selected from 48 submissions. They were organized in topical sections named: Dialogue and Spoken Language Understanding; Language Analysis and Generation; Speech Analysis and Synthesis; Speech Recognition; Text Analysis and Classification.
Offers a wide-ranging overview of the issues and research approaches in the diverse field of applied linguistics Applied linguistics is an interdisciplinary field that identifies, examines, and seeks solutions to real-life language-related issues. Such issues often occur in situations of language contact and technological innovation, where language problems can range from explaining misunderstandings in face-to-face oral conversation to designing automated speech recognition systems for business. The Concise Encyclopedia of Applied Linguistics includes entries on the fundamentals of the discipline, introducing readers to the concepts, research, and methods used by applied linguists working in the field. This succinct, reader-friendly volume offers a collection of entries on a range of language problems and the analytic approaches used to address them. This abridged reference work has been compiled from the most-accessed entries from The Encyclopedia of Applied Linguistics (www.encyclopediaofappliedlinguistics.com), the more extensive volume which is available in print and digital format in 1000 libraries spanning 50 countries worldwide. Alphabetically-organized and updated entries help readers gain an understanding of the essentials of the field with entries on topics such as multilingualism, language policy and planning, language assessment and testing, translation and interpreting, and many others. Accessible for readers who are new to applied linguistics, The Concise Encyclopedia of Applied Linguistics: Includes entries written by experts in a broad range of areas within applied linguistics Explains the theory and research approaches used in the field for analysis of language, language use, and contexts of language use Demonstrates the connections among theory, research, and practice in the study of language issues Provides a perfect starting point for pursuing essential topics in applied linguistics Designed to offer readers an introduction to the range of topics and approaches within the field, The Concise Encyclopedia of Applied Linguistics is ideal for new students of applied linguistics and for researchers in the field.
This two-volume set (CCIS 1393 and CCIS 1394) constitutes selected and revised papers of the 4th International Conference on Advanced Informatics for Computing Research, ICAICR 2020, held in Gurugram, India, in December 2020. The 34 revised full papers and 51 short papers presented were carefully reviewed and selected from 306 submissions. The papers are organized in topical sections on computing methodologies; hardware; networks; security and privacy.
This book constitutes the refereed proceedings of the 7th Iberoamerican Conference on Applications and Usability of Interactive Television, jAUTI 2018, in Bernal, Argentina, in October 2018. The 13 full papers presented were carefully reviewed and selected from numerous submissions. The papers are organized in topical sections on Contexts of application of the IDTV; Design and Implementation Techniques of IDTV Content and Services; Interaction Techniques, Technologies and Accesibility of IDTV Services; Testing and User Experience of IDTV Services.
This volume constitutes the refereed proceedings of the 8th Workshop on Engineering Applications, WEA 2021, held in Medellín, Colombia, in October 2021. Due to the COVID-19 pandemic the conference was held in a hybrid mode. The 33 revised full papers and 11 short papers presented in this volume were carefully reviewed and selected from 127 submissions. The papers are organized in the following topical sections: computational intelligence; bioengineering; Internet of Things (IoT); optimization and operations research; engineering applications.
Sound, devoid of meaning, would not matter to us. It is the information sound conveys that helps the brain to understand its environment. Sound and its underlying meaning are always associated with time and space. There is no sound without spatial properties, and the brain always organizes this information within a temporal–spatial framework. This book is devoted to understanding the importance of meaning for spatial and related further aspects of hearing, including cross-modal inference. People, when exposed to acoustic stimuli, do not react directly to what they hear but rather to what they hear means to them. This semiotic maxim may not always apply, for instance, when the reactions are reflexive. But, where it does apply, it poses a major challenge to the builders of models of the auditory system. Take, for example, an auditory model that is meant to be implemented on a robotic agent for autonomous search-&-rescue actions. Or think of a system that can perform judgments on the sound quality of multimedia-reproduction systems. It becomes immediately clear that such a system needs • Cognitive capabilities, including substantial inherent knowledge • The ability to integrate information across different sensory modalities To realize these functions, the auditory system provides a pair of sensory organs, the two ears, and the means to perform adequate preprocessing of the signals provided by the ears. This is realized in the subcortical parts of the auditory system. In the title of a prior book, the term Binaural Listening is used to indicate a focus on sub-cortical functions. Psychoacoustics and auditory signal processing contribute substantially to this area. The preprocessed signals are then forwarded to the cortical parts of the auditory system where, among other things, recognition, classification, localization, scene analysis, assignment of meaning, quality assessment, and action planning take place. Also, information from different sensory modalities is integrated at this level. Between sub-cortical and cortical regions of the auditory system, numerous feedback loops exist that ultimately support the high complexity and plasticity of the auditory system. The current book concentrates on these cognitive functions. Instead of processing signals, processing symbols is now the predominant modeling task. Substantial contributions to the field draw upon the knowledge acquired by cognitive psychology. The keyword Binaural Understanding in the book title characterizes this shift. Both books, The Technology of Binaural Listening and the current one, have been stimulated and supported by AABBA, an open research group devoted to the development and application of models of binaural hearing. The current book is dedicated to technologies that help explain, facilitate, apply, and support various aspects of binaural understanding. It is organized into five parts, each containing three to six chapters in order to provide a comprehensive overview of this emerging area. Each chapter was thoroughly reviewed by at least two anonymous, external experts. The first part deals with the psychophysical and physiological effects of Forming and Interpreting Aural Objects as well as the underlying models. The fundamental concepts of reflexive and reflective auditory feedback are introduced. Mechanisms of binaural attention and attention switching are covered—as well as how auditory Gestalt rules facilitate binaural understanding. A general blackboard architecture is introduced as an example of how machines can learn to form and interpret aural objects to simulate human cognitive listening. The second part, Configuring and Understanding Aural Space, focuses on the human understanding of complex three-dimensional environments—covering the psychological and biological fundamentals of auditory space formation. This part further addresses the human mechanisms used to process information and interact in complex reverberant environments, such as concert halls and forests, and additionally examines how the auditory system can learn to understand and adapt to these environments. The third part is dedicated to Processing Cross-Modal Inference and highlights the fundamental human mechanisms used to integrate auditory cues with cues from other modalities to localize and form perceptual objects. This part also provides a general framework for understanding how complex multimodal scenes can be simulated and rendered. The fourth part, Evaluating Aural-scene Quality and Speech Understanding, focuses on the object-forming aspects of binaural listening and understanding. It addresses cognitive mechanisms involved in both the understanding of speech and the processing of nonverbal information such as Sound Quality and Quality-of- Experience. The aesthetic judgment of rooms is also discussed in this context. Models that simulate underlying human processes and performance are covered in addition to techniques for rendering virtual environments that can then be used to test these models. The fifth part deals with the Application of Cognitive Mechanisms to Audio Technology. It highlights how cognitive mechanisms can be utilized to create spatial auditory illusions using binaural and other 3D-audio technologies. Further, it covers how cognitive binaural technologies can be applied to improve human performance in auditory displays and to develop new auditory technologies for interactive robots. The book concludes with the application of cognitive binaural technologies to the next generation of hearing aids.
Robust Automatic Speech Recognition: A Bridge to Practical Applications establishes a solid foundation for automatic speech recognition that is robust against acoustic environmental distortion. It provides a thorough overview of classical and modern noise-and reverberation robust techniques that have been developed over the past thirty years, with an emphasis on practical methods that have been proven to be successful and which are likely to be further developed for future applications.The strengths and weaknesses of robustness-enhancing speech recognition techniques are carefully analyzed. The book covers noise-robust techniques designed for acoustic models which are based on both Gaussian mixture models and deep neural networks. In addition, a guide to selecting the best methods for practical applications is provided.The reader will: - Gain a unified, deep and systematic understanding of the state-of-the-art technologies for robust speech recognition - Learn the links and relationship between alternative technologies for robust speech recognition - Be able to use the technology analysis and categorization detailed in the book to guide future technology development - Be able to develop new noise-robust methods in the current era of deep learning for acoustic modeling in speech recognition - The first book that provides a comprehensive review on noise and reverberation robust speech recognition methods in the era of deep neural networks - Connects robust speech recognition techniques to machine learning paradigms with rigorous mathematical treatment - Provides elegant and structural ways to categorize and analyze noise-robust speech recognition techniques - Written by leading researchers who have been actively working on the subject matter in both industrial and academic organizations for many years
This book includes peer-reviewed articles from the 12th International Workshop on Spoken Dialogue System Technology, IWSDS 2021, Singapore. Nowadays, dialogue systems or conversational agents have become one of the most important mechanisms for human-computer or human-robot interaction that has been widely adopted as new paradigm for many applications, companies, and final users. On the other hand, recent advances in natural language processing, understanding and generation, as well as a continuous increasing computational power and large number of resources and data, have brought important and consistent improvements to the capabilities of dialogue systems enabling users to have more productive and enjoyable interactions. However, on the threshold of a new decade, the current state of the art shows important areas where improvements are needed such as incorporation of ground-based knowledge, personality, emotions, and adaptability, as well as automatic mechanisms for objective, robust and fast evaluations, especially in the context of developing social and e-health applications. In this 12th edition of the International Workshop on Spoken Dialogue Systems (IWSDS), “Conversational AI for natural human-centric interaction“ compiles and presents a synopsis on current global research efforts to push forward the state of the art in dialogue technologies, including advances to the classical problems of dialogue management, language generation and understanding, personalisation and generation, spokena and multimodal interaction, dialogue evaluation, dialogue modelling and applications, as well as topics related to chatbots and conversational agent technologies.
