You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to Caffe Users
Hi
I am trying to perform some segmentation task using a CNN in Caffe. For this, I have written a data layer in caffe by myself. The network is receiving some uint8 images as labels. However I checked that the label-image top blob of data layer contains all zeros. Further to ensure that my data layer is reading the label images properly I printed them out to the terminal where the pixel values are being indexed into the top blob of the data layer, and it is printing out non zero values. The entered pixel intensity values range from 0-255 (CVU8C1 format) and I have cast them to Dtype before adding them to the top blob. I am not sure how an Image containing non-zero intensity values is suddenly ending up as all zeros. I am guessing it could have to do with datatype. Can someone give any hint on this?