Is the predicted map the same size as the input image?

leeyeehoo / CSRNet-pytorch

CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes

642 stars 259 forks source link

Is the predicted map the same size as the input image? #43

Open Rayholle opened 5 years ago

Rayholle commented 5 years ago

hello, there is a problem when i run jupyter notebook val.ipynb i try to see the sizes of output and groundtruth and find that the size of output is 1/8 of the input image just like : (704,1024) (88,128) So my problem is Do i need to 8 Bilinear the output of the model before counting it? Thanks a lot

datduyng commented 5 years ago

I have the same question as well. I was wondering if the output image from the model is not the same size as the ground truth GAM then during training how is it able to compute the different. Please advice

Cli98 commented 4 years ago

Yes you should, upsample by factor of 8 and divide pixel by 64