Connectionist Temporal Classification (CTC)

Why we need CTC: RNNs are good for sequential aligned data. The problem is that in speech / handwriting, that is not avaialble.

It is tedious work to create aligned data; so we have CTC loss.

CTC is alignment-free. It works by summing over the probability of all possible alignments between the input and the label.

How it works:

1.) The CTC network assigns the character with the highest probability (or no character) to every input sequence.

2.) Repeats without a blank token in between get merged.

3.) and lastly, the blank token gets removed.

The CTC network can then give you the probability of a label for a given input, by summing over all the probabilities of the character for each time-step.

References: https://distill.pub/2017/ctc/

https://machinelearning-blog.com/2018/09/05/753/