“The largest synthetic dataset for semantic segmentation counting over one million Full HD images taken in 105 different scenarios easily and separately selectable.
Designed to benchmark not only the generalization capability of the Semantic Segmentation architectures but also to assess their domain adaptation performances.”
Overview
Be familiar with IDDA, learn about its peculiarity, the data diversity and the labels description.
Benchmarks
Get in to the challenges opened by our dataset and find out the performances of the treated methods.
