|
|
|
@ -102,7 +102,6 @@ class CTCLabelDecode(BaseRecLabelDecode):
|
|
|
|
|
def __call__(self, preds, label=None, *args, **kwargs):
|
|
|
|
|
if isinstance(preds, paddle.Tensor):
|
|
|
|
|
preds = preds.numpy()
|
|
|
|
|
# out = self.decode_preds(preds)
|
|
|
|
|
|
|
|
|
|
preds_idx = preds.argmax(axis=2)
|
|
|
|
|
preds_prob = preds.max(axis=2)
|
|
|
|
@ -116,27 +115,6 @@ class CTCLabelDecode(BaseRecLabelDecode):
|
|
|
|
|
dict_character = ['blank'] + dict_character
|
|
|
|
|
return dict_character
|
|
|
|
|
|
|
|
|
|
def decode_preds(self, preds):
|
|
|
|
|
probs_ind = np.argmax(preds, axis=2)
|
|
|
|
|
|
|
|
|
|
B, N, _ = preds.shape
|
|
|
|
|
l = np.ones(B).astype(np.int64) * N
|
|
|
|
|
length = paddle.to_tensor(l)
|
|
|
|
|
out = paddle.fluid.layers.ctc_greedy_decoder(preds, 0, length)
|
|
|
|
|
batch_res = [
|
|
|
|
|
x[:idx[0]] for x, idx in zip(out[0].numpy(), out[1].numpy())
|
|
|
|
|
]
|
|
|
|
|
|
|
|
|
|
result_list = []
|
|
|
|
|
for sample_idx, ind, prob in zip(batch_res, probs_ind, preds):
|
|
|
|
|
char_list = [self.character[idx] for idx in sample_idx]
|
|
|
|
|
valid_ind = np.where(ind != 0)[0]
|
|
|
|
|
if len(valid_ind) == 0:
|
|
|
|
|
continue
|
|
|
|
|
conf_list = prob[valid_ind, ind[valid_ind]]
|
|
|
|
|
result_list.append((''.join(char_list), conf_list))
|
|
|
|
|
return result_list
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class AttnLabelDecode(BaseRecLabelDecode):
|
|
|
|
|
""" Convert between text-label and text-index """
|
|
|
|
|