The increasing availability of digital audio content via numerous distribution channels has raised the need for audio retrieval methods that are capable of providing personalized, rapid access to the audio content while dealing with large amounts of audio data. The goal of this special session is to assemble research contributions in the field of audio retrieval, placing emphasis on the intelligence of the proposed techniques. The term intelligence refers to the machine learning properties of the algorithms, their context awareness and the ability to produce high-level metadata from low-level audio descriptors, including content-based summarization, segmentation and clustering descriptions. The special session welcomes contributions on a variety of audio modalities, including speech signals, music sounds and environmental audio data.
Aggelos Pikrakis, Eng., PhD., Lecturer, Dept. of Informatics, University of Piraeus, email@example.com