MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training

• URL: http://arxiv.org/abs/2306.00107v4
• Date: Mon, 22 Apr 2024 21:52:17 GMT
• Title: MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training
• Authors: Yizhi Li, Ruibin Yuan, Ge Zhang, Yinghao Ma, Xingran Chen, Hanzhi Yin, Chenghao Xiao, Chenghua Lin, Anton Ragni, Emmanouil Benetos, Norbert Gyenge, Roger Dannenberg, Ruibo Liu, Wenhu Chen, Gus Xia, Yemin Shi, Wenhao Huang, Zili Wang, Yike Guo, Jie Fu,
• Abstract summary: We propose an acoustic Music undERstanding model with large-scale self-supervised Training (MERT), which incorporates teacher models to provide pseudo labels in the masked language modelling (MLM) style acoustic pre-training. Experimental results indicate that our model can generalise and perform well on 14 music understanding tasks and attain state-of-the-art (SOTA) overall scores.
• Score: 74.32603591331718
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Self-supervised learning (SSL) has recently emerged as a promising paradigm for training generalisable models on large-scale data in the fields of vision, text, and speech. Although SSL has been proven effective in speech and audio, its application to music audio has yet to be thoroughly explored. This is partially due to the distinctive challenges associated with modelling musical knowledge, particularly tonal and pitched characteristics of music. To address this research gap, we propose an acoustic Music undERstanding model with large-scale self-supervised Training (MERT), which incorporates teacher models to provide pseudo labels in the masked language modelling (MLM) style acoustic pre-training. In our exploration, we identified an effective combination of teacher models, which outperforms conventional speech and audio approaches in terms of performance. This combination includes an acoustic teacher based on Residual Vector Quantisation - Variational AutoEncoder (RVQ-VAE) and a musical teacher based on the Constant-Q Transform (CQT). Furthermore, we explore a wide range of settings to overcome the instability in acoustic language model pre-training, which allows our designed paradigm to scale from 95M to 330M parameters. Experimental results indicate that our model can generalise and perform well on 14 music understanding tasks and attain state-of-the-art (SOTA) overall scores.

Related papers

• Self-Supervised Models of Speech Infer Universal Articulatory Kinematics [44.27187669492598]
  We show "inference of articulatory kinematics" as fundamental property of SSL models. We also show that this abstraction is largely overlapping across the language of the data used to train the model. We show that with simple affine transformations, Acoustic-to-Articulatory inversion (AAI) is transferrable across speakers, even across genders, languages, and dialects.
  arXiv  Detail & Related papers  (2023-10-16T19:50:01Z)
• On the Effectiveness of Speech Self-supervised Learning for Music [45.43336822496942]
  Self-sourced learning (SSL) has shown promising results in various speech and natural language processing applications. We explore the music adaption of SSL with two distinctive speech-related models, data2vec1.0 and Hubert, respectively. Our findings suggest that training with music data can generally improve performance on MIR tasks, even when models are trained using paradigms designed for speech.
  arXiv  Detail & Related papers  (2023-07-11T10:37:57Z)
• BEATs: Audio Pre-Training with Acoustic Tokenizers [77.8510930885778]
  Self-supervised learning (SSL) has been witnessed in language, vision, speech, and audio domains over the past few years. We propose BEATs, an iterative audio pre-training framework to learn Bidirectional representation from Audio Transformers. In the first iteration, we use random projection as the acoustic tokenizer to train an audio SSL model in a mask and label prediction manner. Then, we train an acoustic tokenizer for the next iteration by distilling the semantic knowledge from the pre-trained or fine-tuned audio SSL model.
  arXiv  Detail & Related papers  (2022-12-18T10:41:55Z)
• Contrastive Audio-Visual Masked Autoencoder [85.53776628515561]
  Contrastive Audio-Visual Masked Auto-Encoder (CAV-MAE) Our fully self-supervised pretrained CAV-MAE achieves a new SOTA accuracy of 65.9% on VGGSound.
  arXiv  Detail & Related papers  (2022-10-02T07:29:57Z)
• Learning music audio representations via weak language supervision [14.335950077921435]
  We design a multimodal architecture for music and language pre-training (MuLaP) optimised via a set of proxy tasks. weak supervision is provided in the form of noisy natural language descriptions conveying the overall musical content of the track. We demonstrate the usefulness of our approach by comparing the performance of audio representations produced by the same audio backbone with different training strategies.
  arXiv  Detail & Related papers  (2021-12-08T10:30:52Z)
• Contrastive Learning of General-Purpose Audio Representations [33.15189569532155]
  We introduce COLA, a self-supervised pre-training approach for learning a general-purpose representation of audio. We build on recent advances in contrastive learning for computer vision and reinforcement learning to design a lightweight, easy-to-implement model of audio.
  arXiv  Detail & Related papers  (2020-10-21T11:56:22Z)
• Audio ALBERT: A Lite BERT for Self-supervised Learning of Audio Representation [51.37980448183019]
  We propose Audio ALBERT, a lite version of the self-supervised speech representation model. We show that Audio ALBERT is capable of achieving competitive performance with those huge models in the downstream tasks. In probing experiments, we find that the latent representations encode richer information of both phoneme and speaker than that of the last layer.
  arXiv  Detail & Related papers  (2020-05-18T10:42:44Z)
• Curriculum Audiovisual Learning [113.20920928789867]
  We present a flexible audiovisual model that introduces a soft-clustering module as the audio and visual content detector. To ease the difficulty of audiovisual learning, we propose a novel learning strategy that trains the model from simple to complex scene. We show that our localization model significantly outperforms existing methods, based on which we show comparable performance in sound separation without referring external visual supervision.
  arXiv  Detail & Related papers  (2020-01-26T07:08:47Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.