Can caffe input multi-images with just one label?

[Guofeng Hu]

Hi guys,

        

We know caffe is one image to one label, can it input more than one image and they just have one label? Like that (x1,x2,....,x9)-------->y1.  If it can, how to do that? Write a data layer by myself? And how to write it? Thank you

[Przemek D]

Is there any reason you can't simply replicate the label and do x1-->y1,  x2-->y1 ... x9-->y1  ?