I'm trying to predict vector labels for my images, but I'm not sure how to do this in Caffe. For instance, each of my images is associated with a vector of 5 values, in contrast to the examples on the Caffe page, where each image has only 1 value describing the class. Ideally, I would like to create an ImageData layer with input text file:
/path/img1.png 24 1 2.4 45 32
However, the layer documentation doesn't specify if this format is allowed, although I know it can be used for only 1 label. None of the examples on the main site involve vector-valued labels either.
How can I prepare the data or write a prototext definition for the input layer to handle vector labels?