Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
SYSTEM OF GENERATING MODEL FOR ESTIMATION, AND ESTIMATION SYSTEM
Document Type and Number:
WIPO Patent Application WO/2021/005987
Kind Code:
A1
Abstract:
The present invention appropriately performs an estimation based on information that relates to music. A model generation system 10 included in a recommend server 1 generates a model for estimation that is used in estimation based on information relating to music. This system of generating a model for estimation comprises: a learning data acquisition unit 11 for acquiring information for learning which is the learning data used in machine learning and in which information relating to the sound data of a plurality of musical pieces having a sequential order are arranged in order of the sequential order and in the dimension of the sequential order; and a model generation unit 12 for performing machine learning using the information for learning as input to a model for estimation that performs convolution by a kernel in the dimension of the sequential order and generating a model for estimation.

Inventors:
TANAKA SHIGEKI (JP)
FUKAZAWA YUSUKE (JP)
Application Number:
PCT/JP2020/023824
Publication Date:
January 14, 2021
Filing Date:
June 17, 2020
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
NTT DOCOMO INC (JP)
International Classes:
G06N3/08; G06N20/00
Other References:
HSU, KAI-CHUN ET AL.: "Neural Network Based Next- Song Recommendation", ARXIV, 2016, XP055784180, Retrieved from the Internet [retrieved on 20200824]
OORD, AARON VAN DEN, SANDER DIELEMAN; AND BENJAMIN SCHRAUWEN: "Deep content-based music recommendation", ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 26 (NIPS 2013, vol. 26, 2013, pages 2645 - 2653, XP009511517, Retrieved from the Internet [retrieved on 20200824]
Attorney, Agent or Firm:
HASEGAWA Yoshiki et al. (JP)
Download PDF: