Make-An-Audio 2: Temporal-Enhanced Text-to-Audio Generation
This work addresses audio quality and alignment problems in text-to-audio synthesis, which is important for applications like media production and accessibility, but it is incremental as it builds on prior diffusion models.
The paper tackles issues of semantic misalignment and poor temporal consistency in text-to-audio generation by proposing Make-An-Audio 2, which uses LLMs for text parsing and data augmentation, a structured-text encoder, and a Transformer-based denoiser, achieving significant gains in temporal understanding, semantic consistency, and sound quality over baselines.
Large diffusion models have been successful in text-to-audio (T2A) synthesis tasks, but they often suffer from common issues such as semantic misalignment and poor temporal consistency due to limited natural language understanding and data scarcity. Additionally, 2D spatial structures widely used in T2A works lead to unsatisfactory audio quality when generating variable-length audio samples since they do not adequately prioritize temporal information. To address these challenges, we propose Make-an-Audio 2, a latent diffusion-based T2A method that builds on the success of Make-an-Audio. Our approach includes several techniques to improve semantic alignment and temporal consistency: Firstly, we use pre-trained large language models (LLMs) to parse the text into structured <event & order> pairs for better temporal information capture. We also introduce another structured-text encoder to aid in learning semantic alignment during the diffusion denoising process. To improve the performance of variable length generation and enhance the temporal information extraction, we design a feed-forward Transformer-based diffusion denoiser. Finally, we use LLMs to augment and transform a large amount of audio-label data into audio-text datasets to alleviate the problem of scarcity of temporal data. Extensive experiments show that our method outperforms baseline models in both objective and subjective metrics, and achieves significant gains in temporal information understanding, semantic consistency, and sound quality.