speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge pull request #941 from alibaba-damo-academy/dev_cmz

Browse Source 
change eng punc in offline model
This commit is contained in:
chenmengzheAAA 2023-09-12 22:22:06 +08:00 committed by GitHub
commit 60d78d9d84
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 21 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
funasr/bin/punc_infer.py
View File

@ -117,12 +117,25 @@ class Text2Punc:
new_mini_sentence_punc += [int(x) for x in punctuations_np] new_mini_sentence_punc += [int(x) for x in punctuations_np]
words_with_punc = [] words_with_punc = []
for i in range(len(mini_sentence)): for i in range(len(mini_sentence)):
if (i==0 or self.punc_list[punctuations[i-1]] == "。" or self.punc_list[punctuations[i-1]] == "？") and len(mini_sentence[i][0].encode()) == 1:
mini_sentence[i] = mini_sentence[i].capitalize()
if i == 0:
if len(mini_sentence[i][0].encode()) == 1:
mini_sentence[i] = " " + mini_sentence[i]
if i > 0: if i > 0:
if len(mini_sentence[i][0].encode()) == 1 and len(mini_sentence[i - 1][0].encode()) == 1: if len(mini_sentence[i][0].encode()) == 1 and len(mini_sentence[i - 1][0].encode()) == 1:
mini_sentence[i] = " " + mini_sentence[i] mini_sentence[i] = " " + mini_sentence[i]
words_with_punc.append(mini_sentence[i]) words_with_punc.append(mini_sentence[i])
if self.punc_list[punctuations[i]] != "_": if self.punc_list[punctuations[i]] != "_":
words_with_punc.append(self.punc_list[punctuations[i]]) punc_res = self.punc_list[punctuations[i]]
if len(mini_sentence[i][0].encode()) == 1:
if punc_res == "，":
punc_res = ","
elif punc_res == "。":
punc_res = "."
elif punc_res == "？":
punc_res = "?"
words_with_punc.append(punc_res)
new_mini_sentence += "".join(words_with_punc) new_mini_sentence += "".join(words_with_punc)
# Add Period for the end of the sentence # Add Period for the end of the sentence
new_mini_sentence_out = new_mini_sentence new_mini_sentence_out = new_mini_sentence
@ -131,9 +144,15 @@ class Text2Punc:
if new_mini_sentence[-1] == "，" or new_mini_sentence[-1] == "、": if new_mini_sentence[-1] == "，" or new_mini_sentence[-1] == "、":
new_mini_sentence_out = new_mini_sentence[:-1] + "。" new_mini_sentence_out = new_mini_sentence[:-1] + "。"
new_mini_sentence_punc_out = new_mini_sentence_punc[:-1] + [self.period] new_mini_sentence_punc_out = new_mini_sentence_punc[:-1] + [self.period]
elif new_mini_sentence[-1] != "。" and new_mini_sentence[-1] != "？": elif new_mini_sentence[-1] == ",":
new_mini_sentence_out = new_mini_sentence[:-1] + "."
new_mini_sentence_punc_out = new_mini_sentence_punc[:-1] + [self.period]
elif new_mini_sentence[-1] != "。" and new_mini_sentence[-1] != "？" and len(new_mini_sentence[-1].encode())==0:
new_mini_sentence_out = new_mini_sentence + "。" new_mini_sentence_out = new_mini_sentence + "。"
new_mini_sentence_punc_out = new_mini_sentence_punc[:-1] + [self.period] new_mini_sentence_punc_out = new_mini_sentence_punc[:-1] + [self.period]
elif new_mini_sentence[-1] != "." and new_mini_sentence[-1] != "?" and len(new_mini_sentence[-1].encode())==1:
new_mini_sentence_out = new_mini_sentence + "."
new_mini_sentence_punc_out = new_mini_sentence_punc[:-1] + [self.period]
return new_mini_sentence_out, new_mini_sentence_punc_out return new_mini_sentence_out, new_mini_sentence_punc_out