Music Generation Based On Generative Adversarial Networks With Transformer

Abstract

Autoregressive models based on Transformers have become the prevailing approach for generating music compositions that exhibit comprehensive musical structure. These models are typically trained by minimizing the negative log-likelihood (NLL) of the observed sequence in an autoregressive manner. However, when generating long sequences, the quality of samples from these models tends to significantly deteriorate due to exposure bias. To address this issue, we leverage classifiers trained to differentiate between real and sampled sequences to identify these failures. This observation motivates our exploration of adversarial losses as a complement to the NLL objective. We employ a pre-trained Span-BERT model as the discriminator in the Generative Adversarial Network (GAN) framework, which enhances training stability in our experiments. To optimize discrete sequences within the GAN framework, we utilize the Gumbel-Softmax trick to obtain a differentiable approximation of the sampling proces

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyjiang2023music

Related papers

• Gansynth: Adversarial Neural Audio Synthesis (2019)0.00
• Polyphonic Music Generation With Sequence Generative Adversarial Networks (2017)2.26
• Midi-sandwich: Multi-model Multi-task Hierarchical Conditional VAE-GAN Networks For Symbolic Single-track Music Generation (2019)0.00
• Interpretable Melody Generation From Lyrics With Discrete-valued Adversarial Training (2022)6.34
• Objective-reinforced Generative Adversarial Networks (ORGAN) For Sequence Generation Models (2017)0.00
• High Fidelity Speech Synthesis With Adversarial Networks (2019)0.00
• Adversarial Generation Of Time-frequency Features With Application In Audio Synthesis (2019)0.00
• Bandwidth Extension On Raw Audio Via Generative Adversarial Networks (2019)0.00