cardwing / Codes-for-Lane-Detection

Learning Lightweight Lane Detection CNNs by Self Attention Distillation (ICCV 2019)
MIT License
1.04k stars 333 forks source link

How can we test the model using CPU device? #323

Closed Chirag1997aggarwal closed 2 years ago

Chirag1997aggarwal commented 3 years ago

If testing with a CPU device is possible, How can we perform that? Is resizing of images is required for using the data from Tusimple test dataset?

I have tried below command - python tools/test_lanenet.py --weights_path /mnt/usmidet/projects/FORD-DAT/8-Users/GT_Validation_Team/Chirag/CGQ-2485/Codes-for-Lane-Detection-master/weights/model_culane-71-3/culane_lanenet_vgg_2018-12-01-14-38-37.ckpt-10000 --image_path /mnt/usmidet/projects/FORD-DAT/8-Users/GT_Validation_Team/Chirag/CGQ-2485/Codes-for-Lane-Detection-master/img.txt --save_dir /mnt/usmidet/projects/FORD-DAT/8-Users/GT_Validation_Team/Chirag/CGQ-2485/Codes-for-Lane-Detection-master/output_pre/ --use_gpu 0

get the below output on terminal - image

As you can see I have not got any error on terminal but yet not found the output images in output folder.

Kindly help.

cardwing commented 3 years ago

It seems that the code works normally. Can you go to places where variable save_dir is used and check the details? I think the bug is not difficult to fix.