cnexah

Results 9 comments of cnexah

raise ValueError('Non-used anchor #{} found'.format(aind))

I think you can remove some anchors, such as removing some kind of aspect ratio

multi-gpu training

请问multi-gpu会对mono_depth的训练产生影响吗？

replace vgg with resnet

Hi, I also tried Resnet18 and Resnet50, but got a similar performance to VGG.

replace vgg with resnet

Sorry, I made a mistake. My performance of Resnet50 is also bad. I use the **mean absolute error**, and **only test for 'Car'** Angle error, Dimension error VGG19 0.108 [0.086...

replace vgg with resnet

Thanks very much!

replace vgg with resnet

@CaoBo1996 Sorry that I have lost my code. In memory I didn't change much. Maybe just more units in middle layers.

Are DTU results reported in the paper trained on 640 or 1280?

Same question: is training on 1600 x1200 necessary? or we just train on 640 x 512 and test on 1600 x1200?

Are DTU results reported in the paper trained on 640 or 1280?

Thank you!

lac + ganet

Would it be possible to provide the number of parameters of Lac+GANet?