SIR is a benchmarking tool developed in Python, with the help of Torch, built to compare different neureal network (UNET, RES-UNET, DEEPLAB) on the same Computer vision task: semantic segmentation. The dataset used to compare them is https://www.cityscapes-dataset.com/dataset-overview/.
TheNor01/SIR
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|