A Simple Hierarchical Pooling Data Structure for Loop Closure
This work addresses efficiency issues in loop closure for robotics or SLAM systems, but it is incremental as it builds on existing bag-of-word methods with a simpler approach.
The paper tackles the problem of slow loop closure in large-scale applications by proposing a hierarchical pooling data structure that averages bag-of-word descriptors, achieving speedups of 4 to 20 times on benchmark datasets with minimal performance loss.
We propose a data structure obtained by hierarchically averaging bag-of-word descriptors during a sequence of views that achieves average speedups in large-scale loop closure applications ranging from 4 to 20 times on benchmark datasets. Although simple, the method works as well as sophisticated agglomerative schemes at a fraction of the cost with minimal loss of performance.