Re:Deep learning
by Houshalter on Sunday July 26, 2015 @08:57PM
(#9952838)
CNNs aren't rotation invariant on purpose. If they were, you would lose information about how features are oriented. Typically it learns features that are very sensitive to rotation, like vertical edges and horizontal edges.
[ Parent
| on HN ]
I am rather thinking about an on/off 'object detected 'for objects with any rotation angle. That symmetry surely must be exploitable somehow in shared parameters of the DNN or similar?
My gut feeling is that the first convolutional layer's kernels, for example, would probably have a 'some are orthogonal' constraint due to this symmetry.