Kaidi Cao

I am a second-year master student at Computer Science Department, Stanford University. I am working in the Stanford Vision and Learning Lab, under the supervision of Prof. Juan Carlos Niebles and Prof. Li Fei-Fei. I also work with Prof. Tengyu Ma towards some machine learning topics.

I graduated from Tsinghua University in July 2018. Prior to that, I have been lucky to work with Cheng Li(Sensetime) and Prof. Chen Change Loy (NTU) during my internship at Sensetime Research, as well as Dr. Jing Liao and Dr. Lu Yuan when I was a research intern at Visual Computing Group of Microsoft Research Lab - Asia.

My research interests lie primarily in the area of Computer Vision and Machine Learning. To be specific, I'm pariticularly interested in Learning Algorithms and Video Analysis.

Email / LinkedIn / Google Scholar / Github


Sept. 2018 - Jul. 2020 (Expected), Computer Science Department, Stanford University ,

Master of Science, Computer Science.


Aug. 2014 - Jul. 2018 , Department of Electronic Engineering, Tsinghua University ,

Bachelor of Engineering, Electronic Information.


Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss
Kaidi Cao, Colin Wei, Adrien Gaidon, Nikos Arechiga, Tengyu Ma
Neural Information Processing Systems (NeurIPS), 2019
Oral presentation at the Bay Area Machine Learning Symposium
[pdf] [code]

We design two novel methods to improve imbalanced training.


Few-Shot Video Classification via Temporal Alignment
Kaidi Cao, Jingwei Ji*, Zhangjie Cao*, Chien-Yi Chang, Juan Carlos Niebles
Technical report, arXiv:1906.11415, 2019

We propose a video few-shot learning framework that explicitly leverages the temporal ordering information in video data through temporal alignment.


Learning Temporal Action Proposals with Fewer Labels
Jingwei Ji, Kaidi Cao, Juan Carlos Niebles
International Conference on Computer Vision (ICCV), 2019

We propose a semi-supervised learning algorithm for generating temporal action proposals.


Delving Deep into Hybrid Annotations for 3D Human Recovery in the Wild
Yu Rong, Ziwei Liu, Cheng Li, Kaidi Cao, Chen Change Loy
International Conference on Computer Vision (ICCV), 2019
[pdf] [project page] [code]

We provided sufficient investigation of annotation design for in-the-wild 3D human reconstruction.


Geometry-Aware Unsupervised Image-to-Image Translation
Wayne Wu, Kaidi Cao, Cheng Li, Chen Qian, Chen Change Loy
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019
[pdf] [project page]

We propose a geometry-aware framework for unsupervised image-to-image translation, which is robust to arbitrary shape variations between domains.


Disentangling Content and Style via Unsupervised Geometry Distillation
Wayne Wu, Kaidi Cao, Cheng Li, Chen Qian, Chen Change Loy
ICLR Workshop on Deep Generative Models for Highly Structured Data (ICLRW), 2019

We present a framework to learn to disentangle the object into content and style in a completely unsupervised manner.


CariGANs: Unpaired Photo-to-Caricature Translation
Kaidi Cao, Jing Liao, Lu Yuan
ACM Transactions on Graphics, (Proc. of Siggraph Asia), 2018
[pdf] [project page]

We present the first deep learning-based approach to automatically generate the facial caricature for a given portrait photo.

Press Coverage:

Pose-Robust Face Recognition via Deep Residual Equivariant Mapping
Kaidi Cao*, Yu Rong*, Cheng Li, Xiaoou Tang, Chen Change Loy
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018
[pdf][project page][code]

We presented a Deep Residual EquivAriant Mapping (DREAM) block to improve the performance of face recognition on profile faces.


Merge or Not? Learning to Group Faces via Imitation Learning
Yue He*, Kaidi Cao*, Cheng Li, Chen Change Loy
AAAI Conference on Artificial Intelligence (AAAI, Spotlight), 2018

We proposed a novel face grouping framework that makes sequential merging decision based on short- and long-term rewards via inverse reinforcement learning.


DNN Dataflow Choice Is Overrated
Xuan Yang, Mingyu Gao, Jing Pu, Ankita Nayak, Qiaoyi Liu, Steven Emberton Bell, Jeff Ou Setter, Kaidi Cao, Heonjae Ha, Christos Kozyrakis, Mark Horowitz
Technical report, arXiv:1809.04070, 2018

We develop an auto-optimizer for DNN that is closely coupled with the Halide framework.

Academic Services
  • Conference Reviewer: ICCV 2019, AAAI 2020, CVPR 2020

  • Journal Reviewer: IJCV 2018, TNNLS 2019, PAMI 2019