Publications

<*> means equal contribution

2024

  1. ICLR
    Consistency trajectory models: Learning probability flow ode trajectory of diffusion
    Dongjun Kim* , Chieh-Hsin Lai*, Wei-Hsiang Liao , Naoki Murata , Yuhta Takida , Toshimitsu Uesaka , Yutong He , Yuki Mitsufuji , and Stefano Ermon
    International Conference on Learning Representations, 2024
    HTML Blog Code
  2. ICLR
    Manifold preserving guided diffusion
    Yutong He* , Naoki Murata* , Chieh-Hsin Lai, Yuhta Takida , Toshimitsu Uesaka , Dongjun Kim , Wei-Hsiang Liao , Yuki Mitsufuji , J Zico Kolter , Ruslan Salakhutdinov , and Stefano Ermon
    International Conference on Learning Representations, 2024
    HTML Blog
  3. ICLR
    SAN: Inducing metrizability of GAN with discriminative normalized linear layer
    Yuhta Takida , Masaaki Imaizumi , Takashi Shibuya , Chieh-Hsin Lai, Toshimitsu Uesaka , Naoki Murata , and Yuki Mitsufuji
    International Conference on Learning Representations, 2024
    HTML Blog Code
  4. ICASSP
    Vrdmg: Vocal restoration via diffusion posterior sampling with multiple guidance
    Carlos Hernandez-Olivan , Koichi Saito , Naoki Murata , Chieh-Hsin Lai, Marco A Martı́nez-Ramirez , Wei-Hsiang Liao , and Yuki Mitsufuji
    IEEE International Conference on Acoustics, Speech and Signal Processing, 2024
    PDF Blog
  5. TISMIR
    The Sound Demixing Challenge 2023 – Music Demixing Track
    Giorgio Fabbro , Stefan Uhlich , Chieh-Hsin Lai, Woosung Choi , Marco Martı́nez-Ramı́rez , Weihsiang Liao , Igor Gadelha , Geraldo Ramos , Eddie Hsu , Hugo Rodrigues , and  others
    Transactions of the International Society for Music Information Retrieval, 2024
    PDF

2023

  1. ICML
    FP-diffusion: Improving score-based diffusion models by enforcing the underlying score fokker-planck equation
    Chieh-Hsin Lai, Yuhta Takida , Naoki Murata , Toshimitsu Uesaka , Yuki Mitsufuji , and Stefano Ermon
    International Conference on Machine Learning, 2023
    PDF Code
  2. ICML (Oral)
    Gibbsddrm: A partially collapsed gibbs sampler for solving blind inverse problems with denoising diffusion restoration
    Naoki Murata , Koichi Saito , Chieh-Hsin Lai, Yuhta Takida , Toshimitsu Uesaka , Yuki Mitsufuji , and Stefano Ermon
    International Conference on Machine Learning, 2023
    PDF Code
  3. ICASSP
    Unsupervised vocal dereverberation with diffusion-based generative models
    Koichi Saito , Naoki Murata , Toshimitsu Uesaka , Chieh-Hsin Lai, Yuhta Takida , Takao Fukui , and Yuki Mitsufuji
    IEEE International Conference on Acoustics, Speech and Signal Processing, 2023
    PDF Blog
  4. Preprint
    HQ-VAE: Hierarchical Discrete Representation Learning with Variational Bayes
    Yuhta Takida , Yukara Ikemiya , Takashi Shibuya , Kazuki Shimada , Woosung Choi , Chieh-Hsin Lai, Naoki Murata , Toshimitsu Uesaka , Kengo Uchida , Wei-Hsiang Liao , and Yuki Mitsufuji
    arXiv preprint arXiv:2401.00365, 2023
    PDF
  5. Preprint
    On the Language Encoder of Contrastive Cross-modal Models
    Mengjie Zhao , Junya Ono , Zhi Zhong , Chieh-Hsin Lai, Yuhta Takida , Naoki Murata , Wei-Hsiang Liao , Takashi Shibuya , Hiromi Wakaki , and Yuki Mitsufuji
    arXiv preprint arXiv:2310.13267, 2023
    PDF
  6. ICML WS
    On the Equivalence of Consistency-Type Models: Consistency Models, Consistent Diffusion Models, and Fokker-Planck Regularization
    Chieh-Hsin Lai, Yuhta Takida , Toshimitsu Uesaka , Naoki Murata , Yuki Mitsufuji , and Stefano Ermon
    International Conference on Machine Learning 2023 workshop on Structured Probabilistic Inference and Generative Modeling, 2023
    PDF
  7. AISTATS
    Robust variational autoencoding with wasserstein penalty for novelty detection
    Chieh-Hsin Lai*, Dongmian Zou* , and Gilad Lerman
    International Conference on Artificial Intelligence and Statistics, 2023
    PDF

2022

  1. Neurocomputing
    Preventing oversmoothing in VAE via generalized variance parameterization
    Yuhta Takida , Wei-Hsiang Liao , Chieh-Hsin Lai, Toshimitsu Uesaka , Shusuke Takahashi , and Yuki Mitsufuji
    Neurocomputing, 2022
    PDF
  2. ICML
    SQ-VAE: Variational bayes on discrete representation with self-annealed stochastic quantization
    Yuhta Takida , Takashi Shibuya , WeiHsiang Liao , Chieh-Hsin Lai, Junki Ohmura , Toshimitsu Uesaka , Naoki Murata , Shusuke Takahashi , Toshiyuki Kumakura , and Yuki Mitsufuji
    International Conference on Machine Learning, 2022
    PDF Code

2021

  1. Journal
    Breaking symmetries in data-driven phase retrieval
    Raunak Manekar , Kshitij Tayal , Zhong Zhuang , Chieh-Hsin Lai, Vipin Kumar , and Ju Sun
    Computational Optical Sensing and Imaging, 2021
    PDF

2020

  1. ICML WS
    Inverse Problems, Deep Learning, and Symmetry Breaking
    Kshitij Tayal , Chieh-Hsin Lai, Raunak Manekar , Zhong Zhuang , Vipin Kumar , and Ju Sun
    International Conference on Machine Learning 2020 workshop on ML Interpretability for Scientific Discovery, 2020
    PDF
  2. ICLR
    Robust subspace recovery layer for unsupervised anomaly detection
    Chieh-Hsin Lai*, Dongmian Zou* , and Gilad Lerman
    International Conference on Learning Representations, 2020
    HTML PDF Code
  3. NeurIPS WS
    Unlocking inverse problems using deep learning: Breaking symmetries in phase retrieval
    Kshitij Tayal , Chieh-Hsin Lai, Raunak Manekar , Zhong Zhuang , Vipin Kumar , and Ju Sun
    Conference on Neural Information Processing Systems 2020 Workshop on Deep Learning and Inverse Problems, 2020
    PDF