When I follow the tutorial to run the alignment,
for N in {1..4}; do python align_dataset_mtcnn.py ~/datasets/lfw/raw ~/datasets/lfw/lfw_mtcnnpy_160 --image_size 160 --margin 32 --random_order --gpu_memory_fraction 0.25 & done
I got the final result as follows:
Total number of images: 13233
Number of successfully aligned images: 3463
I am doubt that it is the right result. If wrong, what is the right result?
Try to execute the following: python align_dataset_mtcnn.py ~/datasets/lfw/raw ~/datasets/lfw/lfw_mtcnnpy_160 --image_size 160 --margin 32 --random_order --gpu_memory_fraction 0.25 and check the output. The output is confusing because you have executed 4 processes in parallel.
thank you, it works! @dkozlov
Hi guys,
I wonder could you align whole 13233 images? In my case, there are a lot of images unable to be aligned, causing missing pairs when evaluation. I obtain 12784/13233 images.
Most helpful comment
Try to execute the following:
python align_dataset_mtcnn.py ~/datasets/lfw/raw ~/datasets/lfw/lfw_mtcnnpy_160 --image_size 160 --margin 32 --random_order --gpu_memory_fraction 0.25and check the output. The output is confusing because you have executed 4 processes in parallel.