The class converts the enumeration values into one-hot encoding.
void SetEnumSize(int enumSize);
The layer has no trainable parameters.
The single input accepts a blob with int
or float
data that contains the enumeration values, of the dimensions:
Channels
is equal to1
- the other dimensions may be of any size
The single output contains a blob with the vectors that one-hot encode the enumeration values. The dimensions of the blob are:
Channels
is equal toGetEnumSize()
- the other dimensions are the same as for the input