Published October 17, 2022 | Version v2
Dataset Open

Model weights for DiffRoll

Authors/Creators

  • 1. Singapore University of Technology and Design and Sony Group Corporation
  • 2. Sony Group Corporation
  • 3. Singapore University of Technology and Design

Description

This repository contains the pre-trained model weights for DiffRoll proposed in the paper DiffRoll: Diffusion-based Generative Music Transcription with Unsupervised Pretraining Capability by Sony.
More information about DiffRoll including its Demo is available in https://sony.github.io/DiffRoll/.

Files

Files (2.1 GB)

Name Size
md5:472ce6ad7bf6d7329d018e20e8915d79
1.0 GB Download
md5:aca11b73b7eec14f3848a121c6e8673d
1.0 GB Download