Main image created using Microsoft Copilot.
In recent years, the intersection of artificial intelligence (AI), machine learning (ML), and acoustic signal processing has emerged as a rapidly advancing field, offering new ways to analyse, interpret, and enhance sound. The integration of AI and ML technologies into acoustic signal processing is poised to revolutionize a wide range of applications, from speech recognition and audio content analysis to environmental sound monitoring and biomedical diagnostics. This Special Collection (AI and Machine Learning in Acoustic Signal Processing) seeks to capture the latest innovations and breakthroughs in this dynamic area, providing a platform for researchers and practitioners to showcase their work on cutting-edge advancements and challenges.
The primary aim of this Collection is to explore how AI and ML techniques can be harnessed to address complex challenges in acoustic signal processing. These technologies offer significant advantages, such as the ability to process large volumes of data, adapt to new patterns in real-time, and enhance the accuracy and efficiency of signal analysis and enhancement tasks. This Collection aims to attract papers that advance both theoretical research and practical solutions, offering novel tools that can be implemented in real-world applications. The collection will encompass a broad range of topics, including:
- Acoustic Scene Analysis: Developing algorithms for recognizing and classifying various acoustic environments, crucial for applications like smart surveillance, urban noise monitoring, and context-aware systems.
- Signal Enhancement: AI-driven techniques for noise reduction, dereverberation, and signal clarity improvement are of particular interest. These methods are vital in scenarios such as telecommunications, hearing aids, robust speech/underwater communication, and weak underwater signal detection in various types of noisy environments.
- Source Localization and Separation: Advanced ML models for detecting, localizing, and isolating sound sources in multi-source environments. Applications include speaker separation in conference systems, enhancing audio quality in entertainment, improving environmental awareness in autonomous systems, and improving the accuracy and resolution in underwater source localization/detection.
- Speech and Audio Processing: Applying deep learning models to tasks such as robust speech recognition (ASR), audio generation and synthesis, voice conversion, emotion recognition, speech translation. These tasks are essential for developing more natural and intuitive human-computer interaction systems.
- Music Signal Processing: Developing AI models for accessing, analyzing, manipulating, and creating music, such as music information retrieval, music generation, music synthesis, computer accompaniment and machine musicianship.
- Environmental and Biomedical Acoustics: AI applications in these areas involve using acoustic signals for wildlife monitoring, detecting environmental changes, or medical diagnostics, such as analyzing heart or lung sounds. For example processing underwater acoustics signals for marine biology and environmental monitoring, analyzing heart or lung sounds for diagnosis, and ultrasonic signal processing for medical imaging.
- Acoustic Inspection Methods: Novel methodologies for inspection of other assets including the processing and classification of ultrasonic signals. For example, the detection of anomalies in acoustic data, time-series classification, and crack/defect localisation.
Contributions showcasing interdisciplinary approaches or collaborations between AI, ML, and acoustic signal processing experts are particularly encouraged. Submissions that highlight new methodologies, innovative applications, and emerging trends shaping the future of acoustic signal processing are especially welcome.
