Cubical Ripser: Software for computing persistent homology of image and volume data

About

We introduce Cubical Ripser for computing persistent homology of image and volume data (more precisely, weighted cubical complexes). To our best knowledge, Cubical Ripser is currently the fastest and the most memory-efficient program for computing persistent homology of weighted cubical complexes. We demonstrate our software with an example of image analysis in which persistent homology and convolutional neural networks are successfully combined. Our open-source implementation is available online.

Shizuo Kaji, Takeki Sudo, Kazushi Ahara · 2020

Related benchmarks

Task	Dataset	Result
V-filtration computation	Synthetic 2D medium (316 x 316)	Execution Time (s)0.04	5
V-filtration computation	Lena 256 x 256	Execution Time (s)0.02	5
V-filtration computation	Synthetic 2D large 1024 x 1024	Execution Time (s)0.34	5
V-filtration computation	DIV2K 1024 x 1024	Execution Time (s)0.27	5
V-filtration computation	Synthetic 3D medium (46^3)	Execution Time (s)0.31	3
V-filtration computation	Fuel 64^3	Execution Time (s)0.28	3
V-filtration computation	Synthetic 3D large (128^3)	Time (s)8.29	3
V-filtration computation	Bonsai 128^3	Time (s)2.94	3
V-filtration computation	Aneurism 128^3	Time (s)2.62	3