Abstract: Suno has attracted wide attention due to its impressive capabilities. It demonstrates technological advancements and opens up new possibilities for music composition, representing a milestone in the development of artificial intelligence (AI) music generation. In this paper, we first introduce the background and summarize the general technical framework of AI music generation, followed by an analysis of Suno’s advantages and disadvantages. Finally, we discuss the future trends in Music and AI.

Suno：潜力、前景与趋势

[1]Agostinelli A, Denk TI, Borsos Z, et al., 2023. MusicLM: generating music from text. https://arxiv.org/abs/2301.11325

[2]Al-Rfou R, Choe D, Constant N, et al., 2019. Character-level language modeling with deeper self-attention. 33^rd AAAI Conf on Artificial Intelligence, p.3159-3166.

[3]Ao JY, Wang R, Zhou L, et al., 2022. SpeechT5: unified-modal encoder-decoder pre-training for spoken language processing. Proc 60^th Annual Meeting of the Association for Computational Linguistics, p.5723-5738.

[4]Brown TB, Mann B, Ryder N, et al., 2020. Language models are few-shot learners. Proc 34^th Int Conf on Neural Information Processing Systems, Article 159.

[5]Coldewey D, 2022. Try Riffusion, an AI Model That Composes Music by Visualizing It. https://techcrunch.com/2022/12/15/try-riffusion-an-ai-model-that-composes-music-by-visualizing-it/ [Accessed on Apr. 6, 2024].

[6]Copet J, Kreuk F, Gat I, et al., 2023. Simple and controllable music generation. Proc 37^th Int Conf on Neural Information Processing Systems, Article 2066.

[7]Dai ZH, Yang ZL, Yang YM, et al., 2019. Transformer-XL: attentive language models beyond a fixed-length context. Proc 57^th Conf of the Association for Computational Linguistics, p.2978-2988.

[8]Dhariwal P, Jun H, Payne C, et al., 2020. Jukebox: a generative model for music. https://arxiv.org/abs/2005.00341

[9]Freyberg K, 2024. Introducing v3. https://www.suno.ai/blog/v3 [Accessed on Apr. 6, 2024].

[10]Hochreiter S, Schmidhuber J, 1997. Long short-term memory. Neur Comput, 9(8):1735-1780.

[11]Hsiao WY, Liu JY, Yeh YC, et al., 2021. Compound Word Transformer: learning to compose full-song music over dynamic directed hypergraphs. 35^th AAAI Conf on Artificial Intelligence, p.178-186.

[12]Huang CZA, Vaswani A, Uszkoreit J, et al., 2019. Music Transformer: generating music with long-term structure. 7^th Int Conf on Learning Representations.

[13]Huang QQ, Park DS, Wang T, et al., 2023. Noise2Music: text-conditioned music generation with diffusion models. https://arxiv.org/abs/2302.03917

[14]Huang YS, Yang YH, 2020. Pop Music Transformer: beat-based modeling and generation of expressive pop piano compositions. Proc 28^th ACM Int Conf on Multimedia, p.1180-1188.

[15]Kreuk F, Synnaeve G, Polyak A, et al., 2023. AudioGen: textually guided audio generation. 11^th Int Conf on Learning Representations.

[16]Liu HH, Chen ZH, Yuan Y, et al., 2023. AudioLDM: text-to-audio generation with latent diffusion models. Proc 40^th Int Conf on Machine Learning, p.21450-21474.

[17]O’Boyle M, 2023. (Re)Discovering Music Theory: AI Algorithm Learns the Rules of Musical Composition and Provides a Framework for Knowledge Discovery. https://csl.illinois.edu/news-and-media/rediscovering-music-theory-ai-algorithm-learns-the-rules-of-musical-composition-and-provides-a-framework-for-knowledge-discovery [Accessed on Apr. 6, 2024].

[18]Ouyang L, Wu J, Jiang X, et al., 2022. Training language models to follow instructions with human feedback. Proc 36^th Int Conf on Neural Information Processing Systems, Article 2011.

[19]Ren Y, He JZ, Tan X, et al., 2020. PopMAG: pop music accompaniment generation. Proc 28^th ACM Int Conf on Multimedia, p.1198-1206.

[20]Ren Y, Hu CX, Tan X, et al., 2021. FastSpeech 2: fast and high-quality end-to-end text to speech. 9^th Int Conf on Learning Representations.

[21]Touvron H, Martin L, Stone K, et al., 2023. Llama 2: open foundation and fine-tuned chat models. https://arxiv.org/abs/2307.09288

[22]Vaswani A, Shazeer N, Parmar N, et al., 2017. Attention is all you need. Proc 31^st Int Conf on Neural Information Processing Systems, p.6000-6010.

[23]Wu J, Liu XG, Hu XL, et al., 2020. PopMNet: generating structured pop music melodies using neural networks. Artif Intell, 286:103303.

[24]Wu XD, Huang ZJ, Zhang KJ, et al., 2024. MelodyGLM: multi-task pre-training for symbolic melody generation. https://arxiv.org/abs/2309.10738

[25]Yu HZ, Varshney LR, Taube H, et al., 2022. (Re)Discovering laws of music theory using information lattice learning. IEEE BITS Inform Theory Mag, 2(1):58-75.

[26]Yuan RB, Lin HF, Wang Y, et al., 2024. ChatMusician: understanding and generating music intrinsically with LLM. https://arxiv.org/abs/2402.16153

[27]Zeng ML, Tan X, Wang R, et al., 2021. MusicBERT: symbolic music understanding with large-scale pre-training. Findings of the Association for Computational Linguistics, p.791-800.

[28]Zhou J, Ke P, Qiu XP, et al., 2023. ChatGPT: potential, prospects, and limitations. Front Inform Technol Electron Eng, early access.

[29]Zou Y, Zou P, Zhao Y, et al., 2022. MELONS: generating melody with long-term structure using transformers and structure graph. IEEE Int Conf on Acoustics, Speech and Signal Processing, p.191-195.