• Corpus ID: 221319709

The Freesound Loop Dataset and Annotation Tool

  title={The Freesound Loop Dataset and Annotation Tool},
  author={Ant{\'o}nio Ramires and Frederic Font and Dmitry Bogdanov and Jordan B. L. Smith and Yi-Hsuan Yang and Joann Ching and Bo-Yu Chen and Yueh-Kao Wu and Hsu Wei-Han and Xavier Serra},
Music loops are essential ingredients in electronic music production, and there is a high demand for pre-recorded loops in a variety of styles. Several commercial and community databases have been created to meet this demand, but most are not suitable for research due to their strict licensing. We present the Freesound Loop Dataset (FSLD), a new large-scale dataset of music loops annotated by experts. The loops originate from Freesound, a community database of audio recordings released under… 

Figures and Tables from this paper

A Benchmarking Initiative for Audio-Domain Music Generation Using the Freesound Loop Dataset
This paper proposes a new benchmark task for generating musical passages in the audio domain by using the drum loops from the FreeSound Loop Dataset, which are publicly re-distributable, and benchmark the performance of three recent deep generative adversarial network models the authors customize to generate loops, including StyleGAN, StyleGAN2, and UNAGAN.
The results show that a convolutional neural network (CNN), which considers the temporal context of a chord progression, outperforms a simpler approach based on temporal averaging of input features.
Instrument Role Classification: Auto-tagging for Loop Based Music
A new type of auto-tagging task, called “instrument role classification,” is introduced, and the performance of both neural network and non-neural network based multi-label classification models for six instrument roles is benchmarked.


Tonal Description of Polyphonic Audio for Music Content Processing
  • E. Gómez
  • Computer Science
    INFORMS J. Comput.
  • 2006
A method to extract a description of the tonal aspects of music from polyphonic audio signals using different levels of abstraction, differentiating between low-level signal descriptors and high-level textual labels.
Key Estimation in Electronic Dance Music
This paper defines notions of tonality and key before outlining the basic architecture of a template-based key estimation method, and reports on the tonal characteristics of electronic dance music, in order to infer possible modifications of the method described.
Tempo Estimation for Music Loops and a Simple Confidence Measure
Comunicacio presentada a la 17th International Society for Music Information Retrieval Conference (ISMIR 2016), celebrada els dies 7 a 11 d'agost de 2016 a Nova York, EUA.
Freesound technical demo
This demo wants to introduce Freesound to the multimedia community and show its potential as a research resource.
Neural Loop Combiner: Neural Network Models for Assessing the Compatibility of Loops
This work extracts loops from existing music to obtain positive examples of compatible loops, and proposes and compare various strategies for choosing negative examples, and investigates two types of model architectures for estimating the compatibility of loops based on a Siamese network and a pure convolutional neural network.
LoopMaker: Automatic Creation of Music Loops from Pre-recorded Music
LoopMaker is presented, an interactive system that assists users in creating and exploring music loops from pre-recorded music and suggests that the quality of these loops are comparable to manually created loops by experts.
Nonnegative Tensor Factorization for Source Separation of Loops in Audio
This work proposes a novel method of leveraging periodicity in a factorization model: it treats the two-dimensional spectrogram as a three-dimensional tensor, and uses nonnegative tensor factorization to estimate the component spectral templates, rhythms and loop recurrences in a single step.
Finding Drum Breaks in Digital Music Recordings
This paper introduces and formalizes the task of automatically finding suitable drum breaks in music recordings by adapting an approach previously used for singing voice detection, and establishes a first baseline for drum break detection.
Towards Modeling and Decomposing Loop-Based Electronic Music
A prototypical production technique for EM is reviewed and a simplified model is derived from which approaches for the following task are illustrated: given a set of loops that were used to produce a track, decompose the track by finding the points in time at which each loop was activated.
Accurate Tempo Estimation Based on Recurrent Neural Networks and Resonating Comb Filters
A new tempo estimation algorithm which uses a bank of resonating comb filters to determine the dominant periodicity of a musical excerpt by simply reporting the highest resonator’s histogram peak.