ReSync: Riemannian Subgradient-based Robust Rotation Synchronization

Authors: Huikang Liu, Xiao Li, Anthony Man-Cho So

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiment results demonstrate the effectiveness of Re Sync.
Researcher Affiliation Academia Huikang Liu School of Information Management and Engineering Shanghai University of Finance and Economics liuhuikang@shufe.edu.cn; Xiao Li School of Data Science The Chinese University of Hong Kong, Shenzhen lixiao@cuhk.edu.cn; Anthony Man-Cho So Department of Systems Engineering and Engineering Management The Chinese University of Hong Kong manchoso@se.cuhk.edu.hk
Pseudocode Yes Algorithm 1 Re Sync: Riemannian Subgradient Synchronization; Algorithm 2 Spectr In: Spectral Initialization
Open Source Code Yes Our code is available at https://github.com/Huikang2019/Re Sync.
Open Datasets Yes We consider the global alignment problem of three-dimensional scans from the Lucy dataset, which is a down-sampled version of the dataset containing 368 scans with a total number of 3.5 million triangles. We refer to [39] for more details about the experiment setting.
Dataset Splits No The paper describes data generation for synthetic data and the use of the Lucy dataset, but does not specify explicit train/validation/test splits (e.g., percentages, sample counts, or predefined splits) for reproducibility.
Hardware Specification Yes Our experiments are conducted on a personal computer with a 2.90GHz 8-core CPU and 32GB memory.
Software Dependencies No The paper does not provide specific software dependencies with version numbers for its implementation or experiments.
Experiment Setup Yes We use the initial step size µ0 = 1/npq and the decaying factor γ {0.7, 0.8, 0.85, 0.90, 0.95, 0.98} in Re Sync.