miniSAM: A Flexible Factor Graph Non-linear Least Squares Optimization Framework

• DOI: 10.48550/arxiv.1909.00903
• Publication Date: 2019-01-01
• Authors: Jing Dong; Zhaoyang Lv
• Keywords: Python; Benchmarking; Factor graph; Least-squares function approximation; Factor (programming language)

Many problems in computer vision and robotics can be phrased as non-linear least squares optimization represented by factor graphs, for example, simultaneous localization mapping (SLAM), structure from motion (SfM), planning, control. We have developed an open-source C++/Python framework miniSAM, solving such graph based problems. Compared to most existing frameworks solvers, miniSAM has (1) full Python/NumPy API, which enables more agile development easy binding with Python projects, (2) a wide list of sparse linear including CUDA enabled solvers. Our benchmarking results shows offers comparable performances on various types problems, flexible smoother experience.