result with cyclegan

Author: tuanhungvu

README.md

@@ -1,5 +1,8 @@
 # DADA: Depth-aware Domain Adaptation in Semantic Segmentation
 
+## Updates
+- *02/2020*: Using CycleGAN translated images, The DADA model achieves (**43.1%**) on SYNTHIA-2-Cityscapes
+
 ## Paper
 ![](./teaser.png)
 
@@ -99,6 +102,7 @@ For evaluating pretrained networks, execute:
 ```bash
 $ cd DADA/dada/scripts
 $ python test.py --cfg ./<configs_dir>/dada_pretrained.yml
+$ python test.py --cfg ./<configs_dir>/dada_cyclegan_pretrained.yml
 ```
 
 ```<configs_dir>``` could be set as ```configs_s2c``` (```SYNTHIA2Cityscapes```) or ```configs_s2m``` (```SYNTHIA2Mapillary```)

dada/scripts/configs_s2c/dada_cyclegan_pretrained.yml

@@ -0,0 +1,9 @@
+SOURCE: SYNTHIA
+TARGET: Cityscapes
+EXP_NAME: SYNTHIA2Cityscapes_DeepLabv2_DADA_pretrained
+NUM_WORKERS: 4
+USE_DEPTH: True
+TEST:
+  MODE: single
+  RESTORE_FROM:
+    - ../../pretrained_models/synthia2cityscapes_dada_cyclegan.pth