The model was trained on mapillary vistas-dataset. The model was trained for 150k iterations with lr=4e-3, and no wd. You can download the pretrained weights from here Pretrained weights
pgadosey / street-scene-parsing-using-semantic-segmentation Goto Github PK
View Code? Open in Web Editor NEWThis project forked from srihari-humbarwadi/keras-unet
Keras Unet model trained on Mapillary Vistas
License: Apache License 2.0