Content-Based Representations of Audio Using Siamese Neural Networks
- 1 April 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3136-3140
Abstract
In this paper, we focus on the problem of content-based retrieval for audio, which aims to retrieve all semantically similar audio recordings for a given audio clip query. This problem is similar to the problem of query by example of audio, which aims to retrieve media samples from a database, which are similar to the user-provided example. We propose a novel approach which encodes the audio into a vector representation using Siamese Neural Networks. The goal is to obtain an encoding similar for files belonging to the same audio class, thus allowing retrieval of semantically similar audio. Using simple similarity measures such as those based on simple euclidean distance and cosine similarity we show that these representations can be very effectively used for retrieving recordings similar in audio content.Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- Content-based audio classification and retrieval: A novel approach2016
- TUT database for acoustic scene classification and sound event detection2016
- Joint Learning of Speaker and Phonetic Similarities with Siamese Networks2016
- Unsupervised Learning of Visual Representations Using Videos2015
- Detection and Classification of Acoustic Scenes and EventsIEEE Transactions on Multimedia, 2015
- Large-scale content-based audio retrieval from text queries2008
- Content-based multimedia information retrievalACM Transactions on Multimedia Computing, Communications, and Applications, 2006
- Content-based music audio recommendation2005
- Content-based classification, search, and retrieval of audioIEEE MultiMedia, 1996
- Fingerprint enhancement by directional Fourier filteringIEE Proceedings - Vision, Image, and Signal Processing, 1994