A (not so) deep learning model for change detection. The model is implemented to achieve state-of-the-art performance, with a significant reduction in computational complexity and the number of trainable weights compared to the models in the literature. This model can be deployed on edge devices with real-time inference.
This software library contains efficient implementations of Discrete Optimal Transport algorithms for the computation of Kantorovich-Wassestein distances customized for large spatial maps.
The core library is written in standard ANSI-C++11, but it has two wrappers:
A Python wrapper available from PyPI
An R wrapper available from CRAN per EUROSTAT
Optimal transport 1D
The OT1D library offers a simple but efficient implementation of an algorithm to compute the Kantorovich-Wasserstein distance between two empirical measures defined in dimension 1, that is, the support points of the measures are in R. We have designed the algorithm by directly exploiting the Complementary slackness conditions of Linear Programming. The implementation focuses more on efficiency than genericity, and we try to be as efficient as possible in several notable cases. We implemented the core algorithm in standard ANSI C++11, and we provide a python3 wrapper