User profiles for Tianshi Cao
Tianshi CaoPhD student, University of Toronto Verified email at mail.utoronto.ca Cited by 382 |
Texfusion: Synthesizing 3d textures with text-guided image diffusion models
We present TexFusion (Texture Diffusion), a new method to synthesize textures for given 3D
geometries, using only large-scale text-guided image diffusion models. In contrast to recent …
geometries, using only large-scale text-guided image diffusion models. In contrast to recent …
Differentially private diffusion models
While modern machine learning models rely on increasingly large training datasets, data is
often limited in privacy-sensitive domains. Generative models trained with differential privacy …
often limited in privacy-sensitive domains. Generative models trained with differential privacy …
Don't generate me: Training differentially private generative models with sinkhorn divergence
Although machine learning models trained on massive data have led to breakthroughs in
several areas, their deployment in privacy-sensitive domains remains limited due to restricted …
several areas, their deployment in privacy-sensitive domains remains limited due to restricted …
A benchmark of medical out of distribution detection
Motivation: Deep learning models deployed for use on medical tasks can be equipped with
Out-of-Distribution Detection (OoDD) methods in order to avoid erroneous predictions. …
Out-of-Distribution Detection (OoDD) methods in order to avoid erroneous predictions. …
A theoretical analysis of the number of shots in few-shot learning
Few-shot classification is the task of predicting the category of an example from a set of few
labeled examples. The number of labeled examples per category is called the number of …
labeled examples. The number of labeled examples per category is called the number of …
[HTML][HTML] Problems in the deployment of machine-learned models in health care
• Decision-support systems or clinical prediction tools based on machine learning (including
the special case of deep learning) are similar to clinical support tools developed using …
the special case of deep learning) are similar to clinical support tools developed using …
[HTML][HTML] Egocentric video: a new tool for capturing hand use of individuals with spinal cord injury at home
Background Current upper extremity outcome measures for persons with cervical spinal
cord injury (cSCI) lack the ability to directly collect quantitative information in home and …
cord injury (cSCI) lack the ability to directly collect quantitative information in home and …
Scalable neural data server: A data recommender for transfer learning
Absence of large-scale labeled data in the practitioner's target domain can be a bottleneck to
applying machine learning algorithms in practice. Transfer learning is a popular strategy for …
applying machine learning algorithms in practice. Transfer learning is a popular strategy for …
Estimation and analysis of hybrid laminar flow control on a transonic experiment
This paper investigates the e N method for predicting the laminar-to-turbulent transition in
the context of hybrid laminar flow control. Transonic wind tunnel experiments are used to …
the context of hybrid laminar flow control. Transonic wind tunnel experiments are used to …
LATTE3D: Large-scale Amortized Text-To-Enhanced3D Synthesis
Recent text-to-3D generation approaches produce impressive 3D results but require time-consuming
optimization that can take up to an hour per prompt. Amortized methods like ATT3D …
optimization that can take up to an hour per prompt. Amortized methods like ATT3D …