Junyan Jiang (姜峻岩)

One important aspect of human intelligence is the ability to induce patterns and rules from their surroundings without explicit guidance. People utilize such patterns to form concepts which help them conceive higher-level patterns, ultimately leading to a hierarchical understanding of the world. In our study, we want to equip artificial intelligence with a similar ability to understand music, a highly structured sequential data with multiple hierarchies. Our research focuses on the discovery and disentanglement of musically meaningful concepts (e.g., downbeats, chords and phrases) from unlabeled datasets. To achieve this, we adopt various methods to model music syntax and semantics in a self-supervised manner, including Probabilistic Graphical Models (PGMs), Variational Auto-Encoders (VAEs) and contrastive methods. The applications of our research include music generation, music information retrieval and computational musicology across different modality, ranging from music scores, performance controls to audio waveform.

Publications

• Min, L., Jiang, J., Xia, G., Zhao, J. (2023). Polyffusion: A Diffusion Model for Polyphonic Score Generation with Internal and External Controls. ISMIR 2023.
• Jiang, J., Xia, G. (2023). Self-Supervised Hierarchical Metrical Structure Modeling. ICASSP 2023.
• Jiang, J., Chin D., Zhang Y., Xia, G. (2022). Learning Hierarchical Metrical Structure Beyond Measures. ISMIR 2022.
• Zhang Y., Jiang, J., Xia, G., Dixon S. (2022). Interpreting Song Lyrics with an Audio-Informed Pre-trained Language Model. ISMIR 2022.
• Lin, L., Kong, Q., Jiang, J., & Xia, G. (2021). A unified model for zero-shot music source separation, transcription and synthesis. ISMIR 2021.
• Jiang, J., Xia, G., & Berg-Kirkpatrick, T. (2020). Discovering music relations with sequential attention. NLP4MusA workshop in ISMIR 2020.
• Wang, Z., Chen, K., Jiang, J., Zhang, Y., Xu, M., Dai, S., ... & Xia, G. (2020). Pop909: A pop-song dataset for music arrangement generation. ISMIR 2020.
• Wang, Z., Zhang, Y., Zhang, Y., Jiang, J., Yang, R., Zhao, J., & Xia, G. (2020). Pianotree VAE: Structured representation learning for polyphonic music. ISMIR 2020.
• Jiang, J., Xia, G., Carlton, D., Anderson, C., & Miyakawa, R. (2020). Transformer VAE: A Hierarchical Model for Structure-Aware and Interpretable Music Representation Learning. ICASSP 2020.
• Jiang, J., Xia, G., & Dannenberg R. (2019). Representing Music Structure by Variational Attention. ICML 2019 Machine Learning for Music Discovery Workshop.
• Jiang, J., Chen, K., Li, W., & Xia, G. (2019). Large-Vocabulary Chord Transcription via Chord Structure Decomposition. ISMIR 2019.
• Yang, R., Wang, D., Wang, Z., Chen, T., Jiang, J., & Xia, G. (2019). Deep Music Analogy via Latent Representation Disentanglement. ISMIR 2019.
• Jiang, J., Li, W., & Xia, G. (2019). Enhancing Vocal Melody Transcription with Auxiliary Accompaniment Information. CMMR 2019.
• Jiang, J., Chen, K., Li, W., & Xia, G. (2018). MIREX 2018 Submission: A Structural Chord Representation for Automatic Large-Vocabulary Chord Transcription. MIREX extended abstract.
• Jiang, J., Wu, Y., & Li, W. (2017). Extended Abstract for MIREX 2017 Submission: Chord Recognition Using Random Forest Model. MIREX extended abstract.
• Li, P., Hu, G., & Jiang, J. (2017). MIREX 2017 Audio Fingerprinting System. MIREX extended abstract.