A Hypergradient Approach to Robust Regression without Correspondence

Authors: Yujia Xie, Yixiu Mao, Simiao Zuo, Hongteng Xu, Xiaojing Ye, Tuo Zhao, Hongyuan Zha

ICLR 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Thorough numerical experiments show that ROBOT achieves better performance than existing methods in both linear and nonlinear regression tasks, including real-world applications such as flow cytometry and multi-object tracking.
Researcher Affiliation Academia Yujia Xie, Simiao Zuo, Tuo Zhao are affiliated with Georgia Institute of Technology. Emails: {xieyujia, simiaozuo, tourzhao}@gatech.edu. Yixiu Mao is affiliated with Shanghai Jiao Tong University. Email: 956986044myx@gmail.com. Hongteng Xu is affiliated with Gaoling School of Artificial Intelligence, Renmin University of China, and Beijing Key Laboratory of of Big Data Management and Analysis Methods. Email: hongtengxu@ruc.edu.cn. Xiaojing Ye is affiliated with Georgia State University. Email: xye@gsu.edu. Hongyuan Zha is affiliated with School of Data Science, Shenzhen Institute of Artificial Intelligence and Robotics for Society, the Chinese University of Hong Kong, Shenzhen. Email: zhahy@cuhk.edu.cn.
Pseudocode Yes Algorithm 1 Solving S r for robust matching
Open Source Code No The paper mentions obtaining initial weights for a third-party model (Siam RPN) from a GitHub repository, but does not state that the code for their own proposed methodology (ROBOT) is openly available.
Open Datasets Yes For RWOC (i.e., exact correspondence), we use several synthetic regression datasets and a real gated flow cytometry dataset... We adopt the dataset from Knight et al. (2009)... We adopt the MOT17 (Milan et al., 2016) and the MOT20 (Dendorfer et al., 2020) datasets.
Dataset Splits Yes We use 90% of the data as the training data, and the remaining as test data.
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments, such as GPU or CPU models.
Software Dependencies No The paper does not provide specific version numbers for software dependencies or libraries used in the experiments.
Experiment Setup Yes Here, AM and ROBOT is trained with batch size 500 and learning rate 10 4 for 2, 000 iterations. For the Sinkhorn algorithm in ROBOT we set ϵ = 10 4.