Hi, I'm Yanwu Xu (original) (raw)

Short Bio

I am Yanwu Xu, a PhD student in the Electrical and Computer Engineering (ECE) department at Boston University, under the guidance of Dr. Kayhan Batmanghelich and co-advised by Dr. Mingming Gong. Prior to my current endeavors, I earned an M.S. degree from the Intelligent System Program at the University of Pittsburgh in 2022.
My primary research revolves around Generative Models. This encompasses a foundational exploration of Denoising Diffusion models and Generative Adversarial Learning. On the application front, I am engrossed in developing large-scale multimodal generative models for text-to-image generation, relevant in both the computer vision and medical imaging spheres.

Blog

• Congrats on our fundamental generative model accepted by NeurIPS. In this work, we proposed a novel Denoising Diffusion models with GANs.
  Despite the proliferation of generative models, achieving fast sampling during inference without compromising sample diversity and quality remains challenging. Existing models such as Denoising Diffusion Probabilistic Models (DDPM) deliver high-quality, diverse samples but are slowed by an inherently high number of iterative steps. The Denoising Diffusion Generative Adversarial Networks (DDGAN) attempted to circumvent this limitation by integrating a GAN model for larger jumps in the diffusion process. However, DDGAN encountered scalability limitations when applied to large datasets. To address these limitations, we introduce a novel approach that tackles the problem by matching implicit and explicit factors. More specifically, our approach involves utilizing an implicit model to match the marginal distributions of noisy data and the explicit conditional distribution of the forward diffusion. This combination allows us to effectively match the joint denoising distributions. Unlike DDPM but similar to DDGAN, we do not enforce a parametric distribution for the reverse step, enabling us to take large steps during inference. Similar to the DDPM but unlike DDGAN, we take advantage of the exact form of the diffusion process. We demonstrate that our proposed method obtains comparable generative performance to diffusion-based models and vastly superior results to models with a small number of sampling steps.

News

• One paper accepted by NeurIPS2023.
• One paper accepted by ICCV2023.
• One paper accepted by MICCAI2022.
• I will start my 2022 summer internship at Google.
• One papers is accepted by CVPR 2022.
• One papers is accepted by ACM 2021.
• One papers is accepted by ICCV 2021.
• Two papers are accepted by AAAI 2020.
• One paper is accepted by NeurIPS 2019.
• One paper is accepted by BraTS Challenge 2018.
• One paper is accepted by ACCV 2018.