AudioGAN: A Compact and Efficient Framework for Real-Time High-Fidelity Text-to-Audio Generation

Chung H. AudioGAN: A Compact and Efficient Framework for Real-Time High-Fidelity Text-to-Audio Generation. In: 2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio. Audio Engineering Society; 2025.pp. 17:1-10. Available from: https://aes.org/publications/elibrary-page/?id=23006

@inproceedings{Chung2025_23006,
  author = {Chung, Haechun},
  title = {{AudioGAN: A Compact and Efficient Framework for Real-Time High-Fidelity Text-to-Audio Generation}},
  booktitle = {2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio},
  pages = {17:1-10},
  year = {2025},
  month = sep,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23006}
}

TY  - CPAPER
TI  - AudioGAN: A Compact and Efficient Framework for Real-Time High-Fidelity Text-to-Audio Generation
AU  - Chung, Haechun
T2  - 2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio
SP  - 1
EP  - 10
M1  - Paper 17
PY  - 2025
DA  - 2025/09/02
UR  - https://aes.org/publications/elibrary-page/?id=23006
PB  - Audio Engineering Society
LA  - en
AB  - Text-to-audio (TTA) generation can significantly benefit the media industry by reducing production costs and enhancing work efficiency. However, most current TTA modelsprimarily diffusion-based suffer from slow inference speeds and high computational costs. In this paper, we introduce AudioGAN, the first successful GAN-based TTA framework that generates audio in a single pass, thereby reducing model complexity and inference time. To overcome the inherent difficulties in training GANs, we integrate multiple contrastive losses and propose innovative components Single-Double-Triple (SDT) Attention and Time-Frequency Cross-Attention (TF-CA). Extensive experiments on the AudioCaps dataset demonstrate that AudioGAN achieves state-of-the-art performance while using 90% fewer parameters and running 20 times faster, synthesizing audio in under one second. These results establish AudioGAN as a practical and powerful solution for real-time TTA.
ER  -