2026-04-13T02:00:21Zhttps://recercat.cat/oai/request

oai:recercat.cat:10230/454522025-12-20T17:07:05Zcom_2072_6col_2072_452952

urn:hdl:10230/45452 Deep embeddings with Essentia models Alonso-Jiménez, Pablo Bogdanov, Dmitry Serra, Xavier Comunicació presentada a: International Society for Music Information Retrieval Conference celebrat de l'11 al 16 d'octubre de 2020 de manera virtual. We present the integration of various CNN TensorFlow models developed for different MIR tasks into Essentia. This is a continuation of our previous work [1], extending the list of supported models and adding new algorithms to facilitate usability. Essentia provides input feature extraction and inference with TensorFlow models in a single C++ pipeline with Python bindings, facilitating the deployment of C++ and Python MIR applications. We assess the new models’ capabilities to serve as embedding extractors in many downstream classification tasks. All presented models are publicly available on the Essentia website. 2020-10-09T07:34:46Z 2020-10-09T07:34:46Z 2020 info:eu-repo/semantics/conferenceObject info:eu-repo/semantics/acceptedVersion Licensed under a Creative Commons Attribution 4.0 In- ternational License (CC BY 4.0). 21st International Society for Music Information Retrieval Conference, Montréal, Canada, 2020. https://creativecommons.org/licenses/by/4.0/ info:eu-repo/semantics/openAccess ISMIR