This paper studies tagging and retrieval of room impulse responses from a labelled library. A similarity-based method is introduced that relies on perceptually relevant characteristics of reverberation. This method is developed using a publicly available dataset of algorithmic reverberation settings. Semantic word vectors are introduced to exploit semantic correlation among tags and allow for unseen words to be used for retrieval. Average precision is reported on a subset of the dataset as well as tagging of recorded room impulse responses. The developed approach manages to assign downloaded room impulse responses to tags that match their short descriptions. Furthermore, introducing semantic word vectors allows it to perform well even when large portions of the training data have been replaced by synonyms.
Authors:
Chourdakis, Emmanouil Theofanis; Reiss, Joshua D.
Affiliation:
Queen Mary University London, London, UK
AES Convention:
146 (March 2019)
Paper Number:
10198
Publication Date:
March 10, 2019
Subject:
MIR
Click to purchase paper as a non-member or you can login as an AES member to see more options.
No AES members have commented on this paper yet.
To be notified of new comments on this paper you can subscribe to this RSS feed. Forum users should login to see additional options.
If you are not yet an AES member and have something important to say about this paper then we urge you to join the AES today and make your voice heard. You can join online today by clicking here.