Zhifeng Kong

University of California, San Diego

I am a PhD in Computer Science at University of California San Diego (UCSD). My advisor is Professor Kamalika Chaudhuri. I work on deep generative models, including Diffusion Model, GAN, Normalizing Flow, and VAE. Specifically, I build state-of-the-art generative (diffusion) models for different applications. Besides diffusion models, I conduct research to understand deep generative models, including their expressivity, reliability, controllability, and trustworthiness. I am also widely interested in conditional and multi-modal generative models. I hold a B.S. in mathematics and applied mathematics from Xi'an Jiaotong University, China, in which I was luckily enrolled in the Honors Youth Program (少年班, 2012-2014) and the National Honors Science Program (珠峰计划, 2014-2018). My Erdős number is 3.

Email: z4kong at eng dot ucsd dot edu

Website Google Scholar CV

News (June 2023)

• I defended on May 25, 2023. My doctorate dissertation is about expressivity and trustworthiness of deep generative models.

• I will join NVIDIA as a research scientist in July, 2023.

Research

Papers on Deep Generative Models

• Zhifeng Kong, Kamalika Chaudhuri. Data Redaction from Conditional Generative Models.
[paper] [Tag: Conditional Generative Models, Trustworthiness]

• Zhifeng Kong, Kamalika Chaudhuri. Data Redaction from Pre-trained GANs. In SaTML 2023.
[paper] [video] [Tag: GAN, Trustworthiness]

• Zhifeng Kong, Scott Alfeld. Approximate Data Deletion in Generative Models.
[paper] [Tag: Generative Model, Privacy, Trustworthiness]

• Zhifeng Kong, Kamalika Chaudhuri. Understanding Instance-based Interpretability of Variational Auto-Encoders. In NeurIPS 2021.
[paper] [code] [video] [blog] [Tag: VAE, Interpretability]

• Zhaoyang Lyu*, Zhifeng Kong*, Xudong Xu, Liang Pan, Dahua Lin. A conditional point diffusion-refinement paradigm for 3d point cloud completion. In ICLR 2022.
[paper] [code] [Tag: Diffusion Model, Point Cloud Completion]

• Zhifeng Kong, Wei Ping. On Fast Sampling of Diffusion Probabilistic Models. In ICML 2021 INNF Workshop.
[paper] [demo] [code] [Tag: Diffusion Model, Sampling]

• Zhifeng Kong, Wei Ping, Jiaji Huang, Kexin Zhao, Bryan Catanzaro. DiffWave: A Versatile Diffusion Model for Audio Synthesis. In ICLR 2021 (oral).
[paper] [demo] [video] [code part 1] [code part 2] [Tag: Diffusion Model, Speech Synthesis]

• Zhifeng Kong, Kamalika Chaudhuri. Universal Approximation of Residual Flows in Maximum Mean Discrepancy. In ICML 2021 INNF Workshop.
[paper] [video] [Tag: Normalizing Flows, Theory]

• Zhifeng Kong, Kamalika Chaudhuri. The Expressive Power of a Class of Normalizing Flow Models. In AISTATS 2020.
[paper] [slides] [blog] [video (workshop version)] [Tag: Normalizing Flows, Theory]

Papers on Speech Enhancement

• Zhifeng Kong, Wei Ping, Ambrish Dantrey, Bryan Catanzaro. Speech Denoising in the Waveform Domain with Self-Attention. In ICASSP 2022.
[paper] [demo] [code] [Tag: Self Attention, Speech Denoising]

• Zhifeng Kong, Wei Ping, Ambrish Dantrey, Bryan Catanzaro. CleanUNet 2: A Hybrid Speech Denoising Model on Waveform and Spectrogram. In INTERSPEECH 2023.
[Tag: Self Attention, Speech Denoising]

Papers on Trustworthy ML (Supervised Learning)

• Zhifeng Kong*, Amrita Roy Chowdhury*, Kamalika Chaudhuri. Can Membership Inferencing be Refuted?
[paper] [Tag: Membership Inference]

• Zhaoyang Lyu, Ching-Yun Ko, Zhifeng Kong, Ngai Wong, Dahua Lin, Luca Daniel. Fastened CROWN: Tightened Neural Network Robustness Certificates. In AAAI 2020.
[paper] [Tag: Robustness]

Blogs and Technical Reports

• Understanding Instance-based Interpretability of Variational Auto-Encoders. [blog]

• The expressive power of a class of normalizing flow models. [blog]

• Generalization theory for GANs. [slides]

• Notes on generalization properties of GAN metrics. [pdf]

• Invertible ResNets. [slides]

• Zhihu Blogs

• UCSD Machine Learning Blogs

Industrial Experiences

• Research Internship at Baidu Research, USA. (Summer 2020)

• Research Internship at NVIDIA Applied Deep Learning Research. (Summer 2021 & 2022)

Services

I served as reviewers for JMLR, ICML 2023, ICLR 2023, AISTATS 2023, NeurIPS 2022, ICLR 2022, ICML 2022, ECCV 2022 Workshop UNCV, UAI 2022 Workshop TPM, NeurIPS 2021, ICML 2020 WorkShop WHI. I was an outstanding reviewer at ICML 2022.