About
https://research.nvidia.com/person/can-zhao
https://scholar.google.com/citations?user=C…
Activity
-
Congratulations! Great collaboration work
Congratulations! Great collaboration work
Liked by Can Zhao
-
Meet many friends old and new, and somewhere in-between at the AIMBE 2026 annual meeting.
Meet many friends old and new, and somewhere in-between at the AIMBE 2026 annual meeting.
Liked by Can Zhao
-
The Rosie challenge brings together amazing work by MSOE students every year, pushing the boundary of what it means to develop AI as an I undergrad…
The Rosie challenge brings together amazing work by MSOE students every year, pushing the boundary of what it means to develop AI as an I undergrad…
Liked by Can Zhao
Experience
-
-
-
-
-
Education
-
-
-
-
Licenses & Certifications
Publications
-
MAISI-v2: Accelerated 3D High-Resolution Medical Image Synthesis with Rectified Flow and Region-specific Contrastive Loss
AAAI 2026
See publicationMAISI-v1 (DDPM) and MAISI-v2 (Flow matching) are the first 3D medical image generation Foundation Models that can work for flexible body region, voxel size, and volume size. Please check our GUI demo https://lnkd.in/ggvEkV8E
and paper https://lnkd.in/gyp58NUg
-
iSMORE: An Iterative Self Super-Resolution Algorithm
Best paper at MICCAI SASHIMI workshop
See publicationThe first self super-resolution algorithm that enables 3D CNN
-
Applications of a deep learning method for anti-aliasing and super-resolution in MRI
Magnetic resonance imaging
See publicationThis paper reviews our MRI self super-resolution algorithm SMORE and then demonstrates its performance in four applications with the goal to demonstrate its potential for use in both research and clinical scenarios.
-
A Deep Learning based Anti-aliasing Self Super-resolution Algorithm for Magnetic Resonance Imaging
International Conference on Medical Image Computing and Computer-Assisted Intervention
See publicationDeep network based super-resolution and anti-aliasing for MRI without any external training data
-
Self super-resolution for magnetic resonance images using deep networks
International Symposium on Biomedical Imaging
See publicationDeep network based super-resolution for MRI without any external training data
-
A Supervoxel Based Random Forest Synthesis Framework for Bidirectional MR/CT Synthesis
International Conference on Medical Image Computing and Computer-Assisted Intervention, International Workshop on Simulation and Synthesis in Medical Imaging
See publicationCT/MR image synthesis using random forest and super-voxel clustering.
-
Whole Brain Segmentation and Labeling from CT Using Synthetic MR Images
International Conference on Medical Image Computing and Computer-Assisted Intervention, International Workshop on Machine Learning in Medical Imaging
See publicationCT to MR image synthesis using deep network,
whole brain segmentation on CT images. -
Effects of spatial resolution on image registration
SPIE Medical Imaging
See publicationTheoretically explored effects of image spatial resolution on image registration.
More activity by Can
-
It took great effort to secure a new location for MICCAI 2026 just 5 months before the conference. I’m happy to announce that we’ll now gather in…
It took great effort to secure a new location for MICCAI 2026 just 5 months before the conference. I’m happy to announce that we’ll now gather in…
Liked by Can Zhao
-
🚀 Call for Participation | MRIxFields2026 We are excited to launch another MICCAI challenge: MRIxFields2026, focusing on cross-field MRI generation…
🚀 Call for Participation | MRIxFields2026 We are excited to launch another MICCAI challenge: MRIxFields2026, focusing on cross-field MRI generation…
Liked by Can Zhao
-
Excited to be at NVIDIA GTC 2026 this week! Healthcare AI is rapidly evolving—from foundation models to Physical AI systems that can understand and…
Excited to be at NVIDIA GTC 2026 this week! Healthcare AI is rapidly evolving—from foundation models to Physical AI systems that can understand and…
Liked by Can Zhao
-
AI isn’t a single model. It’s a full stack. Energy. Chips. Infrastructure. Models. Applications. That’s the five-layer cake powering the largest…
AI isn’t a single model. It’s a full stack. Energy. Chips. Infrastructure. Models. Applications. That’s the five-layer cake powering the largest…
Liked by Can Zhao
-
I’m grateful and excited to share that I’ve stepped into the role of 𝐃𝐢𝐫𝐞𝐜𝐭𝐨𝐫 𝐨𝐟 𝐭𝐡𝐞 𝐃𝐢𝐯𝐢𝐬𝐢𝐨𝐧 𝐨𝐟 𝐀𝐈 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 at…
I’m grateful and excited to share that I’ve stepped into the role of 𝐃𝐢𝐫𝐞𝐜𝐭𝐨𝐫 𝐨𝐟 𝐭𝐡𝐞 𝐃𝐢𝐯𝐢𝐬𝐢𝐨𝐧 𝐨𝐟 𝐀𝐈 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 at…
Liked by Can Zhao
-
Excited to attend NVIDIA GTC 2026! Looking forward to connecting with researchers, engineers, and innovators pushing the boundaries of AI…
Excited to attend NVIDIA GTC 2026! Looking forward to connecting with researchers, engineers, and innovators pushing the boundaries of AI…
Liked by Can Zhao
-
Not to be missed if you work on AV, Robotics, or Machine Intelligence.
Not to be missed if you work on AV, Robotics, or Machine Intelligence.
Liked by Can Zhao
-
We are excited to release the weights of COLIPRI, our 3D vision–language encoder for chest CT scans, on Hugging Face 🤗 As you can see in the…
We are excited to release the weights of COLIPRI, our 3D vision–language encoder for chest CT scans, on Hugging Face 🤗 As you can see in the…
Liked by Can Zhao
-
⭐ In the latest issue of the #MICCAI Quarterly newsletter, Marius George Linguraru said, "As we reach the end of my first year as president of our…
⭐ In the latest issue of the #MICCAI Quarterly newsletter, Marius George Linguraru said, "As we reach the end of my first year as president of our…
Liked by Can Zhao
-
I’m excited to present our work MAISI-v2 on AAAI 2026. MAISI-v1 (DDPM) and MAISI-v2 (Flow matching) are the first 3D medical image generation…
I’m excited to present our work MAISI-v2 on AAAI 2026. MAISI-v1 (DDPM) and MAISI-v2 (Flow matching) are the first 3D medical image generation…
Shared by Can Zhao
Other similar profiles
Explore more posts
-
EGK Microelectronic Solutions Group Sdn. Bhd.
Large Language Models (LLMs), Support Vector Machines (SVM), Decision Trees, Random Forests, and Gradient Boosting are among the most powerful tools in the modern machine learning (ML) toolbox. While LLMs have recently gained significant attention in mainstream media due to their transformative capabilities in natural language processing and AI applications, traditional ML models like SVM, Decision Trees, and ensemble methods such as Random Forests and Gradient Boosting remain crucial in solving structured data problems across engineering disciplines. Yet, it is surprising how many engineers still have not integrated machine learning into their workflows to enhance productivity, optimize design, or automate repetitive tasks. As industries continue to move toward digital transformation and Industry 4.0, the adoption of ML is no longer optional—it is a competitive advantage. #LLM, #SVM, #DecisionTree #RandomForests #GradientBoosting #EGK #IsaacLab #Industry40
-
Anyscale
Physical Intelligence turns terabytes (TBs) of robotics sensor and video data into high-value training datasets. Before Ray on Anyscale, scaling multimodal data pipelines across diverse hardware slowed progress and demanded deep infrastructure expertise, pulling developers away from differentiated research work. With Anyscale, Physical Intelligence enables more than a dozen of its developers to run experiments concurrently and deploy workloads seamlessly across a multi-cloud environment, accelerating iteration and keeping teams focused on their mission to build advanced AI for robotics. Read the full story: https://lnkd.in/g29uj4vi
-
PyTorch
PyTorch 2.8 brings native XCCL support to Intel® GPUs, enabling seamless distributed training. Case studies on Argonne’s Aurora supercomputer show 99% scaling efficiency and Llama3 pre-training at scale. 🔗 Read our latest blog from Intel Corporation's PyTorch Team & Argonne National Laboratory: https://hubs.la/Q03JbBcf0 #PyTorch #Intel #AIInfrastructure #OpenSourceAI
1 Comment -
Ke Zou
We’re excited to share our latest work: “Percolative Pathway to Stripe Order in KTaO₃-Based Superconductivity” is now available on arXiv! https://lnkd.in/geiKXEDs In this study, we show that controlled interfacial disorder in MgO/KTaO₃(111) heterostructures triggers a novel superconducting transition, evolving from localized Cooper-pair islands to superconducting puddles and eventually to stripe-ordered superconductivity. Importantly, the observed stripe width matches the spin precession length, suggesting a self-organized modulation driven by spin–orbit coupling and symmetry breaking.
2 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content