tf.keras.ops.ctc_decode

Decodes the output of a CTC model.

tf.keras.ops.ctc_decode(
    inputs,
    sequence_lengths,
    strategy='greedy',
    beam_width=100,
    top_paths=1,
    merge_repeated=True,
    mask_index=None
)

inputs A tensor of shape (batch_size, max_length, num_classes) containing the logits (the output of the model). They should not be normalized via softmax.
sequence_lengths A tensor of shape (batch_size,) containing the sequence lengths for the batch.
strategy A string for the decoding strategy. Supported values are "greedy" and "beam_search".
beam_width An integer scalar beam width used in beam search. Defaults to 100.
top_paths An integer scalar, the number of top paths to return. Defaults to 1.
merge_repeated A boolean scalar, whether to merge repeated labels in the output. Defaults to True.
mask_index An integer scalar, the index of the mask character in the vocabulary. Defaults to None.

A tuple containing:

  • The tensor representing the list of decoded sequences. If strategy="greedy", the shape is (1, batch_size, max_length). If strategy="beam_seatch", the shape is (top_paths, batch_size, max_length). Note that: -1 indicates the blank label.
  • If strategy="greedy", a tensor of shape (batch_size, 1) representing the negative of the sum of the probability logits for each sequence. If strategy="beam_seatch", a tensor of shape (batch_size, top_paths) representing the log probability for each sequence.