About Me

I’m Siyuan Shan (单思远) with a Ph.D. degree in Computer Science obtained from UNC-Chapel Hill, working with Prof. Junier Oliva. My research interests lie in machine learning, especially speech and music genAI, sequence modeling, time series imputation, meta learning, and set modeling.

I received my bachelor’s degree from the Department of Biomedical Engineering, Beihang University in 2017. During undergraduate studying, I was working with Prof. Yan Xu, focusing on medical image registration and video text recognition.

Education

• 2018 - 2022: Ph.D. in Computer Science, The University of North Carolina at Chapel Hill, USA
• 2015 - 2016: Exchange Student, Czech Technical University in Prague, Czech Republic
• 2013 - 2017: B.E. from Department of Biomedical Engineering, Beihang University, China

Experience

• 2023.02 - now: Senior Data Scientist - Conversational AI at Walmart Global Tech.
• 2022.05 - 2022.08: Research Intern at Bytedance, Mountain View. My advisor is Mingbo Ma.
• 2021.05 - 2021.08: Research Intern at Bytedance, Mountain View. My advisor is Hanoi Hantrakul⁣ and Jitong Chen.
• 2020.05 - 2020.08: Research Intern at Bytedance, Mountain View. My advisor is Jitong Chen.
• 2018.03 - 2018.06: Kaggle Freesound General-Purpose Audio Tagging Challenge, 16/558 (top 3%).

Publication

• Siyuan Shan, Yang Li, Amartya Banerjee, Junier B. Oliva. “Phoneme Hallucinator: One-shot Voice Conversion via Set Expansion,” in AAAI Conference on Artificial Intelligence (AAAI) 2024. [Demo, Codes] [Talk]

• Siyuan Shan, Yang Li, Junier Oliva. “NRTSI: Non-Recurrent Time Series Imputation,” in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2023 (Oral). [Talk]

• Siyuan Shan, Vishal Baskaran, Haidong Yi, Jolene Ranek, Natalie Stanley and Junier Oliva. “Transparent Single-Cell Set Classification with Kernel Mean Embeddings,” in ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM-BCB) 2022 (Best Paper Award). [Talk]

• Vishal Baskaran, Jolene Ranek, Siyuan Shan, Natalie Stanley and Junier Oliva. “Distribution-based Sketching of Single-Cell Samples,” in ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM-BCB) 2022.

• Siyuan Shan, Lamtharn Hantrakul, Jitong Chen, Matt Avent, David Trevelyan. “Differentiable Wavetable Synthesis,” in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2022. [Talk] [3rd Party Codes]

• Yang Li, Siyuan Shan, Qin Liu, Junier Oliva. “Towards Robust Active Feature Acquisition,” in Arxiv 2021.

• Siyuan Shan, Yang Li, Junier Oliva. “Meta-Neighborhoods,” in Neural Information Processing Systems (NeurIPS) 2020. [Talk]

• Yang Li, Haidong Yi, Christopher M Bender, Siyuan Shan, Junier Oliva. “Exchangeable Neural ODE for Set Modeling,” in Neural Information Processing Systems (NeurIPS) 2020.

• Meijun Liu, Jicong Zhang, Wenxiao Jia, Qi Chang, Siyuan Shan, Yegang Hu, Dangxiao Wang. “Enhanced executive attention efficiency after adaptive force control training: behavioural and physiological results,” in Behavioural Brain Research 2019.

• Yan Xu, Siyuan Shan, Ziming Qiu, Zhipeng Jia, Zhengyang Shen, Yipei Wang, Mengfei Shi, I Eric, Chao Chang. “End-to-end subtitle detection and recognition for videos in East Asian languages via CNN ensemble,” in Signal Processing: Image Communication 2018.

• Siyuan Shan, Wen Yan, Xiaoqing Guo, Eric I-Chao Chang, Yubo Fan, Yan Xu. “Unsupervised end-to-end learning for deformable medical image registration,” in Arxiv 2017.

Acamedic Service

• Reviewer for ICML 2021-2024, NeurIPS 2021-2023, ICLR 2022-2024, ACM-MM 2022, TCBB, Interspeech 2023-2024, WASPAA 2023, ICASSP 2024.

Patent

• “Differentiable wavetable synthesizer,” in US Patent App. 17/525,814.

MISC

• I was born in Beijing.

• I loves playing soccer and I am a big fan of Lionel Messi and FC Barcelona.