This book provides a comprehensive overview of the latest research breakthroughs and technological applications in modern speech technology, with a special focus on deep learning and multimodal integration. The book explores how advances in machine learning, attention mechanisms, and cross-modal information fusion are reshaping the field. It addresses current limitations in noise robustness, emotion perception, contextual understanding, and human-like speech synthesis, providing innovative solutions through both theoretical frameworks and cutting-edge models. The aim is to offer a unified view of how speech systems can be enhanced in accuracy, naturalness, and interactivity. By integrating perspectives from linguistics, signal processing, and affective computing, the book appeals to an interdisciplinary audience seeking to understand and shape the future of speech-based intelligence.

Professor Liejun Wang received the PhD degree from Xi'an Jiaotong University in 2012. In 2013, he was appointed as Vice Dean and Associate Professor of the School of Computer Science and Technology at Xinjiang University, where his research focused on wireless sensor theory and the application of key technologies. In 2014, he was promoted to Director of the Network and Information Technology Center and Professor at Xinjiang University, with research focusing on network communication related technologies. In 2016, he became Dean of the School of Software at Xinjiang University. In 2018, he was appointed as both Party Secretary and Dean of the School of Computer Science and Technology at Xinjiang University. In 2024, he was promoted to Vice President of Xinjiang University. In terms of scientific research, supported by the National Science and Technology Major Project and the National Natural Science Foundation of China, he has dedicated his work to pioneering novel theoretical methods and key technologies in multimedia information processing. His achievements include the development of an innovative workshop scheduling optimization framework based on deep reinforcement learning (DRL), which addresses critical limitations of traditional scheduling methods, such as poor adaptability in dynamic environments, multi-objective optimization conflicts, and inefficient real-time decision-making. Furthermore, he has advanced multimedia perception and fusion methodologies, enabling in-depth mining and efficient collaboration of multi-source heterogeneous workshop data. These breakthroughs have positioned his research at the forefront of international scholarship, establishing a theoretical foundation for intelligent, flexible, and efficient scheduling in smart manufacturing workshops—particularly for precision decision-making in complex discrete manufacturing scenarios.

Since graduating, he has published 70 papers indexed by SCI-Expanded and 40 papers indexed by EI. According to the Web of Science Core Collection, his papers have garnered 1,667 citations, and have also received a high number of citations in the Google Scholar database. He was honored with the Second Prize of the Natural Science Award of the Xinjiang Uygur Autonomous Region in 2022 (ranking first), the Medal for Development and Construction of Xinjiang in 2024 (ranking first), and was selected for the Tianshan Talents - High-Level Leading Talent Program of the Xinjiang Uygur Autonomous Region in 2023. In addition, he has completed three monographs, including "Electronic Integrated Design and Experiment" (Xi'an Jiaotong University Press, 2010), "Principles and Applications of Single-Chip Microcomputers" (Xi'an Jiaotong University Press, 2012), and "Industrial Internet Security" (China Machine Press, 2024).