My research journey showcases a diverse portfolio of projects with broad-ranging applications, including:

• Smart infrastructure and urban environments
• Analysis of everyday soundscapes
• Human–machine interaction and assistive technologies
• Acoustic monitoring in healthcare settings
• Intelligent noise monitoring

My involvement in mentoring students at various stages of their academic journey spans a wide range of topics and applications, including:

• Zero-shot learning
• Representation learning
• Active learning
• Deep learning for sound event detection
• Compensation for loudspeaker distortions
• Audio captioning
• Tire type classification
• Semi-supervised musical instrument recognition
• Guitar chord detection
• Parameter adaptation in nonlinear loudspeaker models

Deep Learning for Audio: Applying new approaches for audio content analysis.

• Introduction to computational audio content analysis
• Multi-label zero-shot audio classification, conference publication
• Positive and negative sampling strategies for self-supervised learning, conference publication
• Active learning for sound event detection, journal publication
• Sequential information in polyphonic sound event detection, conference publication

Sound Event Detection (SED): Designing robust models to detect and timestamp sound events in real-world, often noisy, environments. See a detailed introduction to sound event detection in everyday environments.

• Introduction to sound event detection in everyday environments
• Sound event detection with soft labels, conference publication
• Sound event detection: a tutorial, Magazine publication
• Sound event detection in the DCASE 2017 Challenge, journal publication
• Convolutional recurrent neural networks, journal publication

Acoustic Scene Classification (ASC): the process of training computers to identify different environments, such as streets, parks, or offices, based on their distinct acoustic characteristics.

• Data-efficient low-complexity ASC, conference publication
• Low-complexity ASC for multi-device audio, conference publication
• Audio-visual scene analysis, conference publication
• Generalization across devices, conference publication
• Closed and open set classification, conference publication

Evaluation Frameworks: Developing standardized metrics, protocols, and datasets for benchmarking audio analysis systems.

• Evaluation metrics for sound event detection
• Evaluation toolbox for Sound Event Detection
• Datasets and Evaluation, book chapter
• Metrics for polyphonic sound event detection, journal publication
• Joint measurement of localization and detection of sound events, conference publication
• Benchmarks for DCASE Challenge

Dataset Development: Creating annotated audio datasets and evaluation benchmarks for reproducible research, and the release of open datasets.

• Data acquisition: audio collection, annotation, and dataset creation
• Details about the key datasets used in my research
• Open datasets created and maintained by me
• Data-related tool
• DCASE DataList

Real-time Audio Analysis: Implementing real-time systems for on-device and streaming audio recognition.
See details about some related projects:

• Smart cities
• Assistive technologies
• Healthcare environments

Academic Impact: My research centers on neural networks, deep learning, and audio signal processing.

• 60+ publications, with over 8K citations (h-index 35).
• SciVal ranks me 8^th worldwide in scholarly output and 20^th in field-weighted citation impact for "Neural networks; Deep learning; Audio Signal Processing" (2014–2023).

Open Data and Community Resources:

• Co-authored and published over 20 open-access audio datasets at Tampere University, downloaded more than 200K times and used as benchmarks across machine learning competitions and academic studies.
• Maintaining DCASE DataList

Evaluation Standards and Tools: I have contributed to defining standard evaluation metrics for SED and sound event localization and detection (SELD), and helped establish evaluation setups for ASC:

• Metrics for polyphonic sound event detection, journal publication
• Joint measurement of localization and detection of sound events, conference publication

I also maintain several open-source toolboxes that are widely used in the community:

• sed_eval (83K pip downloads, 143 GitHub stars)
• dcase_util (128K pip downloads, 130 GitHub stars)
• sed_vis (120 GitHub stars)

Service to the Research Community:

• Peer reviewer for leading journals and conferences, including IEEE TASLP, IEEE JSTSP, ICASSP, WASPAA, EUSIPCO, DCASE, ISMIR, and AES.
• Organizational roles within the DCASE Challenge (task coordinator, webmaster) and the DCASE Workshop (proceedings editor, technical chair, publication chair).

Academic Illustrations in DCASE Research:

• Numerous illustrations for educational and scientific purposes, visually explaining complex concepts such as sound event detection, acoustic scene classification, and audio tagging.
• Illustrations are provided with Creative Commons license to enhance accessibility and understanding in the field, benefiting researchers, educators, and students.

Academic biography: See a concise overview of my career in academia, highlighting relevant qualifications, experience, and accomplishments.