{ }
View source on GitHub |
Performs beam search decoding on the logits given in input.
tf.compat.v1.nn.ctc_beam_search_decoder(
inputs, sequence_length, beam_width=100, top_paths=1, merge_repeated=True
)
ctc_beam_search_decoder
treats blanks as sequence terminationctc_greedy_decoder
treats blanks as regular elements
If merge_repeated
is True
, merge repeated classes in the output beams.
This means that if consecutive entries in a beam are the same,
only the first of these is emitted. That is, when the sequence is
A B B * B * B
(where '*' is the blank label), the return value is:
A B
ifmerge_repeated = True
.A B B B
ifmerge_repeated = False
.