AES E-Library

Mana, a Tool for Human-Supervised Statistical Analysis in Audio-Content Extraction

Nowadays, meta-data and audio descriptors extraction (used in classification, for instance) is engaged in a rather blind and brute-force method, computing the most possible and then selecting what works using an often long and boring statistical analysis. Moreover, this analysis barely takes into account the intrinsic sense these descriptors may carry. Mana is graphical user-interface (GUI) based system that aims at adding a bit of human supervision in the process, combining state-of-the-art classification methods (in audio-content extraction and classification) with an ease of use that provides the user with direct control over descriptors and their significance in classification.

Author: Wack, Nicolas
Affiliation: Music Technology Group, Institut Universitari de l'Audiovisual, Universitat Pompeu Fabra, Barcelona, Spain
AES Convention:117 (October 2004) Paper Number:6205 Import into BibTeX
Session Subject:Music Synthesis; Audio Archiving, Storage, and Restoration: Content Management

Click to purchase paper or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!

This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.

Learn more about the AES E-Library

E-Library Location: (CD 117Papers)   /117/6205.pdf

Start a discussion about this paper!