Hugging Face's logo

Scicom-intl
/
CallEnhancer

Audio-to-Audio
speech-restoration
speech-enhancement
bandwidth-extension
telephony
call-center
w2v-bert
lora
dac
Model card Files
xet
 Community
huseinzolkepliscicom commited on
Commit
f0bea0e
·
verified ·
1 Parent(s): ea9469d

CallEnhancer v1, re-published without the four Malaysian example calls

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +407 -0
  2. README.md +142 -0
  3. assets/callcentre_cer.png +3 -0
  4. callhome_benchmark/callhome_benchmark.zip +3 -0
  5. callhome_benchmark/results.md +11 -0
  6. decoder_callcentre/decoder_only.pt +3 -0
  7. decoder_callcentre/last.pt +3 -0
  8. decoder_only.pt +3 -0
  9. decoder_train_state.pt +3 -0
  10. examples/callhome_deu.wav +3 -0
  11. examples/callhome_deu_restored48k_step108000.wav +3 -0
  12. examples/callhome_deu_restored48k_step120000.wav +3 -0
  13. examples/callhome_deu_restored48k_step132000.wav +3 -0
  14. examples/callhome_deu_restored48k_step144000.wav +3 -0
  15. examples/callhome_deu_restored48k_step148000.wav +3 -0
  16. examples/callhome_deu_restored48k_step158000.wav +3 -0
  17. examples/callhome_deu_restored48k_step168000.wav +3 -0
  18. examples/callhome_deu_restored48k_step180000.wav +3 -0
  19. examples/callhome_deu_restored48k_step194000.wav +3 -0
  20. examples/callhome_deu_restored48k_step212000.wav +3 -0
  21. examples/callhome_deu_restored48k_step226000.wav +3 -0
  22. examples/callhome_deu_restored48k_step242000.wav +3 -0
  23. examples/callhome_deu_restored48k_step262000.wav +3 -0
  24. examples/callhome_deu_restored48k_step286000.wav +3 -0
  25. examples/callhome_deu_restored48k_step310000.wav +3 -0
  26. examples/callhome_deu_restored48k_step332000.wav +3 -0
  27. examples/callhome_deu_restored48k_step342000.wav +3 -0
  28. examples/callhome_deu_restored48k_step356000.wav +3 -0
  29. examples/callhome_deu_restored48k_step382000.wav +3 -0
  30. examples/callhome_deu_restored48k_step392000.wav +3 -0
  31. examples/callhome_deu_restored48k_step408000.wav +3 -0
  32. examples/callhome_deu_restored48k_step432000.wav +3 -0
  33. examples/callhome_deu_restored48k_step444000.wav +3 -0
  34. examples/callhome_deu_restored48k_step458000.wav +3 -0
  35. examples/callhome_deu_restored48k_step46000.wav +3 -0
  36. examples/callhome_deu_restored48k_step484000.wav +3 -0
  37. examples/callhome_deu_restored48k_step492000.wav +3 -0
  38. examples/callhome_deu_restored48k_step500000.wav +3 -0
  39. examples/callhome_deu_restored48k_step56000.wav +3 -0
  40. examples/callhome_deu_restored48k_step66000.wav +3 -0
  41. examples/callhome_deu_restored48k_step78000.wav +3 -0
  42. examples/callhome_deu_restored48k_step94000.wav +3 -0
  43. examples/callhome_eng.wav +3 -0
  44. examples/callhome_eng_restored48k_step108000.wav +3 -0
  45. examples/callhome_eng_restored48k_step120000.wav +3 -0
  46. examples/callhome_eng_restored48k_step132000.wav +3 -0
  47. examples/callhome_eng_restored48k_step144000.wav +3 -0
  48. examples/callhome_eng_restored48k_step148000.wav +3 -0
  49. examples/callhome_eng_restored48k_step158000.wav +3 -0
  50. examples/callhome_eng_restored48k_step168000.wav +3 -0
.gitattributes CHANGED
@@ -33,3 +33,410 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ examples/audio1.mp3 filter=lfs diff=lfs merge=lfs -text
37
+ examples/callhome_deu.wav filter=lfs diff=lfs merge=lfs -text
38
+ examples/callhome_eng.wav filter=lfs diff=lfs merge=lfs -text
39
+ examples/callhome_jpn.wav filter=lfs diff=lfs merge=lfs -text
40
+ examples/callhome_spa.wav filter=lfs diff=lfs merge=lfs -text
41
+ examples/callhome_zho.wav filter=lfs diff=lfs merge=lfs -text
42
+ examples/callhome_deu_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
43
+ examples/callhome_eng_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
44
+ examples/callhome_jpn_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
45
+ examples/callhome_spa_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
46
+ examples/callhome_zho_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
47
+ examples/malaysian1_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
48
+ examples/malaysian2_restored48k_step46000.wav filter=lfs diff=lfs merge=lfs -text
49
+ examples/callhome_deu_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
50
+ examples/callhome_eng_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
51
+ examples/callhome_jpn_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
52
+ examples/callhome_spa_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
53
+ examples/callhome_zho_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
54
+ examples/malaysian1_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
55
+ examples/malaysian2_restored48k_step56000.wav filter=lfs diff=lfs merge=lfs -text
56
+ examples/callhome_deu_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
57
+ examples/callhome_eng_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
58
+ examples/callhome_jpn_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
59
+ examples/callhome_spa_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
60
+ examples/callhome_zho_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
61
+ examples/malaysian1_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
62
+ examples/malaysian2_restored48k_step66000.wav filter=lfs diff=lfs merge=lfs -text
63
+ examples/callhome_deu_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
64
+ examples/callhome_eng_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
65
+ examples/callhome_jpn_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
66
+ examples/callhome_spa_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
67
+ examples/callhome_zho_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
68
+ examples/malaysian1_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
69
+ examples/malaysian2_restored48k_step78000.wav filter=lfs diff=lfs merge=lfs -text
70
+ examples/callhome_deu_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
71
+ examples/callhome_eng_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
72
+ examples/callhome_jpn_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
73
+ examples/callhome_spa_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
74
+ examples/callhome_zho_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
75
+ examples/malaysian1_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
76
+ examples/malaysian2_restored48k_step94000.wav filter=lfs diff=lfs merge=lfs -text
77
+ examples/callhome_deu_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
78
+ examples/callhome_eng_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
79
+ examples/callhome_jpn_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
80
+ examples/callhome_spa_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
81
+ examples/callhome_zho_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
82
+ examples/malaysian1_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
83
+ examples/malaysian2_restored48k_step108000.wav filter=lfs diff=lfs merge=lfs -text
84
+ examples/callhome_deu_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
85
+ examples/callhome_eng_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
86
+ examples/callhome_jpn_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
87
+ examples/callhome_spa_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
88
+ examples/callhome_zho_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
89
+ examples/malaysian1_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
90
+ examples/malaysian2_restored48k_step120000.wav filter=lfs diff=lfs merge=lfs -text
91
+ examples/callhome_deu_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
92
+ examples/callhome_eng_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
93
+ examples/callhome_jpn_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
94
+ examples/callhome_spa_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
95
+ examples/callhome_zho_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
96
+ examples/malaysian1_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
97
+ examples/malaysian2_restored48k_step132000.wav filter=lfs diff=lfs merge=lfs -text
98
+ examples/callhome_deu_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
99
+ examples/callhome_eng_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
100
+ examples/callhome_jpn_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
101
+ examples/callhome_spa_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
102
+ examples/callhome_zho_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
103
+ examples/malaysian1_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
104
+ examples/malaysian2_restored48k_step144000.wav filter=lfs diff=lfs merge=lfs -text
105
+ examples/callhome_deu_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
106
+ examples/callhome_eng_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
107
+ examples/callhome_jpn_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
108
+ examples/callhome_spa_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
109
+ examples/callhome_zho_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
110
+ examples/malaysian1_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
111
+ examples/malaysian2_restored48k_step148000.wav filter=lfs diff=lfs merge=lfs -text
112
+ examples/callhome_deu_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
113
+ examples/callhome_eng_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
114
+ examples/callhome_jpn_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
115
+ examples/callhome_spa_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
116
+ examples/callhome_zho_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
117
+ examples/malaysian1_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
118
+ examples/malaysian2_restored48k_step158000.wav filter=lfs diff=lfs merge=lfs -text
119
+ examples/callhome_deu_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
120
+ examples/callhome_eng_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
121
+ examples/callhome_jpn_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
122
+ examples/callhome_spa_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
123
+ examples/callhome_zho_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
124
+ examples/malaysian1_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
125
+ examples/malaysian2_restored48k_step168000.wav filter=lfs diff=lfs merge=lfs -text
126
+ examples/callhome_deu_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
127
+ examples/callhome_eng_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
128
+ examples/callhome_jpn_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
129
+ examples/callhome_spa_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
130
+ examples/callhome_zho_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
131
+ examples/malaysian1_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
132
+ examples/malaysian2_restored48k_step180000.wav filter=lfs diff=lfs merge=lfs -text
133
+ examples/callhome_deu_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
134
+ examples/callhome_eng_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
135
+ examples/callhome_jpn_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
136
+ examples/callhome_spa_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
137
+ examples/callhome_zho_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
138
+ examples/malaysian1_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
139
+ examples/malaysian2_restored48k_step194000.wav filter=lfs diff=lfs merge=lfs -text
140
+ examples/callhome_deu_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
141
+ examples/callhome_eng_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
142
+ examples/callhome_jpn_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
143
+ examples/callhome_spa_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
144
+ examples/callhome_zho_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
145
+ examples/malaysian1_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
146
+ examples/malaysian2_restored48k_step212000.wav filter=lfs diff=lfs merge=lfs -text
147
+ examples/callhome_deu_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
148
+ examples/callhome_eng_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
149
+ examples/callhome_jpn_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
150
+ examples/callhome_spa_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
151
+ examples/callhome_zho_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
152
+ examples/malaysian1_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
153
+ examples/malaysian2_restored48k_step226000.wav filter=lfs diff=lfs merge=lfs -text
154
+ examples/callhome_deu_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
155
+ examples/callhome_eng_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
156
+ examples/callhome_jpn_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
157
+ examples/callhome_spa_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
158
+ examples/callhome_zho_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
159
+ examples/malaysian1_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
160
+ examples/malaysian2_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
161
+ examples/malaysian3_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
162
+ examples/malaysian4_restored48k_step242000.wav filter=lfs diff=lfs merge=lfs -text
163
+ examples/callhome_deu_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
164
+ examples/callhome_eng_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
165
+ examples/callhome_jpn_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
166
+ examples/callhome_spa_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
167
+ examples/callhome_zho_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
168
+ examples/malaysian1_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
169
+ examples/malaysian2_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
170
+ examples/malaysian3_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
171
+ examples/malaysian4_restored48k_step262000.wav filter=lfs diff=lfs merge=lfs -text
172
+ examples/callhome_deu_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
173
+ examples/callhome_eng_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
174
+ examples/callhome_jpn_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
175
+ examples/callhome_spa_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
176
+ examples/callhome_zho_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
177
+ examples/malaysian1_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
178
+ examples/malaysian2_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
179
+ examples/malaysian3_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
180
+ examples/malaysian4_restored48k_step286000.wav filter=lfs diff=lfs merge=lfs -text
181
+ examples/callhome_deu_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
182
+ examples/callhome_eng_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
183
+ examples/callhome_jpn_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
184
+ examples/callhome_spa_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
185
+ examples/callhome_zho_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
186
+ examples/malaysian1_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
187
+ examples/malaysian2_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
188
+ examples/malaysian3_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
189
+ examples/malaysian4_restored48k_step310000.wav filter=lfs diff=lfs merge=lfs -text
190
+ examples/callhome_deu_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
191
+ examples/callhome_eng_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
192
+ examples/callhome_jpn_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
193
+ examples/callhome_spa_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
194
+ examples/callhome_zho_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
195
+ examples/malaysian1_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
196
+ examples/malaysian2_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
197
+ examples/malaysian3_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
198
+ examples/malaysian4_restored48k_step332000.wav filter=lfs diff=lfs merge=lfs -text
199
+ examples/callhome_deu_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
200
+ examples/callhome_eng_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
201
+ examples/callhome_jpn_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
202
+ examples/callhome_spa_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
203
+ examples/callhome_zho_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
204
+ examples/malaysian1_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
205
+ examples/malaysian2_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
206
+ examples/malaysian3_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
207
+ examples/malaysian4_restored48k_step342000.wav filter=lfs diff=lfs merge=lfs -text
208
+ examples/callhome_deu_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
209
+ examples/callhome_eng_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
210
+ examples/callhome_jpn_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
211
+ examples/callhome_spa_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
212
+ examples/callhome_zho_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
213
+ examples/malaysian1_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
214
+ examples/malaysian2_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
215
+ examples/malaysian3_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
216
+ examples/malaysian4_restored48k_step356000.wav filter=lfs diff=lfs merge=lfs -text
217
+ examples/callhome_deu_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
218
+ examples/callhome_eng_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
219
+ examples/callhome_jpn_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
220
+ examples/callhome_spa_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
221
+ examples/callhome_zho_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
222
+ examples/malaysian1_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
223
+ examples/malaysian2_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
224
+ examples/malaysian3_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
225
+ examples/malaysian4_restored48k_step382000.wav filter=lfs diff=lfs merge=lfs -text
226
+ examples/callhome_deu_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
227
+ examples/callhome_eng_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
228
+ examples/callhome_jpn_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
229
+ examples/callhome_spa_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
230
+ examples/callhome_zho_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
231
+ examples/malaysian1_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
232
+ examples/malaysian2_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
233
+ examples/malaysian3_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
234
+ examples/malaysian4_restored48k_step392000.wav filter=lfs diff=lfs merge=lfs -text
235
+ examples/callhome_deu_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
236
+ examples/callhome_eng_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
237
+ examples/callhome_jpn_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
238
+ examples/callhome_spa_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
239
+ examples/callhome_zho_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
240
+ examples/malaysian1_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
241
+ examples/malaysian2_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
242
+ examples/malaysian3_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
243
+ examples/malaysian4_restored48k_step408000.wav filter=lfs diff=lfs merge=lfs -text
244
+ examples/callhome_deu_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
245
+ examples/callhome_eng_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
246
+ examples/callhome_jpn_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
247
+ examples/callhome_spa_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
248
+ examples/callhome_zho_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
249
+ examples/malaysian1_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
250
+ examples/malaysian2_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
251
+ examples/malaysian3_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
252
+ examples/malaysian4_restored48k_step432000.wav filter=lfs diff=lfs merge=lfs -text
253
+ examples/callhome_deu_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
254
+ examples/callhome_eng_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
255
+ examples/callhome_jpn_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
256
+ examples/callhome_spa_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
257
+ examples/callhome_zho_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
258
+ examples/malaysian1_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
259
+ examples/malaysian2_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
260
+ examples/malaysian3_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
261
+ examples/malaysian4_restored48k_step444000.wav filter=lfs diff=lfs merge=lfs -text
262
+ examples/callhome_deu_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
263
+ examples/callhome_eng_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
264
+ examples/callhome_jpn_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
265
+ examples/callhome_spa_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
266
+ examples/callhome_zho_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
267
+ examples/malaysian1_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
268
+ examples/malaysian2_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
269
+ examples/malaysian3_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
270
+ examples/malaysian4_restored48k_step458000.wav filter=lfs diff=lfs merge=lfs -text
271
+ examples/callhome_deu_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
272
+ examples/callhome_eng_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
273
+ examples/callhome_jpn_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
274
+ examples/callhome_spa_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
275
+ examples/callhome_zho_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
276
+ examples/malaysian1_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
277
+ examples/malaysian2_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
278
+ examples/malaysian3_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
279
+ examples/malaysian4_restored48k_step484000.wav filter=lfs diff=lfs merge=lfs -text
280
+ examples/callhome_deu_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
281
+ examples/callhome_eng_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
282
+ examples/callhome_jpn_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
283
+ examples/callhome_spa_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
284
+ examples/callhome_zho_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
285
+ examples/malaysian1_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
286
+ examples/malaysian2_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
287
+ examples/malaysian3_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
288
+ examples/malaysian4_restored48k_step492000.wav filter=lfs diff=lfs merge=lfs -text
289
+ examples/callhome_deu_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
290
+ examples/callhome_eng_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
291
+ examples/callhome_jpn_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
292
+ examples/callhome_spa_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
293
+ examples/callhome_zho_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
294
+ examples/malaysian1_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
295
+ examples/malaysian2_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
296
+ examples/malaysian3_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
297
+ examples/malaysian4_restored48k_step500000.wav filter=lfs diff=lfs merge=lfs -text
298
+ assets/callcentre_cer.png filter=lfs diff=lfs merge=lfs -text
299
+ ab_degrade_eb/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_base.wav filter=lfs diff=lfs merge=lfs -text
300
+ ab_degrade_eb/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_clean.wav filter=lfs diff=lfs merge=lfs -text
301
+ ab_degrade_eb/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_hard1.wav filter=lfs diff=lfs merge=lfs -text
302
+ ab_degrade_eb/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_hard2.wav filter=lfs diff=lfs merge=lfs -text
303
+ ab_degrade_eb/01_rainbow_08_highpitch_base.wav filter=lfs diff=lfs merge=lfs -text
304
+ ab_degrade_eb/01_rainbow_08_highpitch_clean.wav filter=lfs diff=lfs merge=lfs -text
305
+ ab_degrade_eb/01_rainbow_08_highpitch_hard1.wav filter=lfs diff=lfs merge=lfs -text
306
+ ab_degrade_eb/01_rainbow_08_highpitch_hard2.wav filter=lfs diff=lfs merge=lfs -text
307
+ ab_degrade_eb/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_base.wav filter=lfs diff=lfs merge=lfs -text
308
+ ab_degrade_eb/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_clean.wav filter=lfs diff=lfs merge=lfs -text
309
+ ab_degrade_eb/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_hard1.wav filter=lfs diff=lfs merge=lfs -text
310
+ ab_degrade_eb/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_hard2.wav filter=lfs diff=lfs merge=lfs -text
311
+ ab_degrade_eb/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_base.wav filter=lfs diff=lfs merge=lfs -text
312
+ ab_degrade_eb/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_clean.wav filter=lfs diff=lfs merge=lfs -text
313
+ ab_degrade_eb/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_hard1.wav filter=lfs diff=lfs merge=lfs -text
314
+ ab_degrade_eb/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_hard2.wav filter=lfs diff=lfs merge=lfs -text
315
+ ab_degrade_eb/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_base.wav filter=lfs diff=lfs merge=lfs -text
316
+ ab_degrade_eb/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_clean.wav filter=lfs diff=lfs merge=lfs -text
317
+ ab_degrade_eb/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_hard1.wav filter=lfs diff=lfs merge=lfs -text
318
+ ab_degrade_eb/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_hard2.wav filter=lfs diff=lfs merge=lfs -text
319
+ ab_degrade_eb/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_base.wav filter=lfs diff=lfs merge=lfs -text
320
+ ab_degrade_eb/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_clean.wav filter=lfs diff=lfs merge=lfs -text
321
+ ab_degrade_eb/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_hard1.wav filter=lfs diff=lfs merge=lfs -text
322
+ ab_degrade_eb/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_hard2.wav filter=lfs diff=lfs merge=lfs -text
323
+ ab_degrade_eb/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_base.wav filter=lfs diff=lfs merge=lfs -text
324
+ ab_degrade_eb/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_clean.wav filter=lfs diff=lfs merge=lfs -text
325
+ ab_degrade_eb/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_hard1.wav filter=lfs diff=lfs merge=lfs -text
326
+ ab_degrade_eb/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_hard2.wav filter=lfs diff=lfs merge=lfs -text
327
+ ab_degrade_eb/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_base.wav filter=lfs diff=lfs merge=lfs -text
328
+ ab_degrade_eb/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_clean.wav filter=lfs diff=lfs merge=lfs -text
329
+ ab_degrade_eb/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_hard1.wav filter=lfs diff=lfs merge=lfs -text
330
+ ab_degrade_eb/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_hard2.wav filter=lfs diff=lfs merge=lfs -text
331
+ ab_degrade_eb/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_base.wav filter=lfs diff=lfs merge=lfs -text
332
+ ab_degrade_eb/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_clean.wav filter=lfs diff=lfs merge=lfs -text
333
+ ab_degrade_eb/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_hard1.wav filter=lfs diff=lfs merge=lfs -text
334
+ ab_degrade_eb/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_hard2.wav filter=lfs diff=lfs merge=lfs -text
335
+ ab_degrade_eb/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_base.wav filter=lfs diff=lfs merge=lfs -text
336
+ ab_degrade_eb/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_clean.wav filter=lfs diff=lfs merge=lfs -text
337
+ ab_degrade_eb/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_hard1.wav filter=lfs diff=lfs merge=lfs -text
338
+ ab_degrade_eb/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_hard2.wav filter=lfs diff=lfs merge=lfs -text
339
+ ab_degrade_eb/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_base.wav filter=lfs diff=lfs merge=lfs -text
340
+ ab_degrade_eb/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_clean.wav filter=lfs diff=lfs merge=lfs -text
341
+ ab_degrade_eb/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_hard1.wav filter=lfs diff=lfs merge=lfs -text
342
+ ab_degrade_eb/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_hard2.wav filter=lfs diff=lfs merge=lfs -text
343
+ ab_degrade_eb/11_rainbow_03_fast_base.wav filter=lfs diff=lfs merge=lfs -text
344
+ ab_degrade_eb/11_rainbow_03_fast_clean.wav filter=lfs diff=lfs merge=lfs -text
345
+ ab_degrade_eb/11_rainbow_03_fast_hard1.wav filter=lfs diff=lfs merge=lfs -text
346
+ ab_degrade_eb/11_rainbow_03_fast_hard2.wav filter=lfs diff=lfs merge=lfs -text
347
+ ab_degrade_eb2/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_base.wav filter=lfs diff=lfs merge=lfs -text
348
+ ab_degrade_eb2/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_clean.wav filter=lfs diff=lfs merge=lfs -text
349
+ ab_degrade_eb2/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_hard1.wav filter=lfs diff=lfs merge=lfs -text
350
+ ab_degrade_eb2/00_\#5[[:space:]]Music,[[:space:]]Life,[[:space:]]and[[:space:]]Love[[:space:]](ft.[[:space:]]SoImJenn[[:space:]]&_hard2.wav filter=lfs diff=lfs merge=lfs -text
351
+ ab_degrade_eb2/00_argentinian_es_000309_base.wav filter=lfs diff=lfs merge=lfs -text
352
+ ab_degrade_eb2/00_argentinian_es_000309_clean.wav filter=lfs diff=lfs merge=lfs -text
353
+ ab_degrade_eb2/00_argentinian_es_000309_hard1.wav filter=lfs diff=lfs merge=lfs -text
354
+ ab_degrade_eb2/00_argentinian_es_000309_hard2.wav filter=lfs diff=lfs merge=lfs -text
355
+ ab_degrade_eb2/01_rainbow_08_highpitch_base.wav filter=lfs diff=lfs merge=lfs -text
356
+ ab_degrade_eb2/01_rainbow_08_highpitch_clean.wav filter=lfs diff=lfs merge=lfs -text
357
+ ab_degrade_eb2/01_rainbow_08_highpitch_hard1.wav filter=lfs diff=lfs merge=lfs -text
358
+ ab_degrade_eb2/01_rainbow_08_highpitch_hard2.wav filter=lfs diff=lfs merge=lfs -text
359
+ ab_degrade_eb2/01_south_african_002577_base.wav filter=lfs diff=lfs merge=lfs -text
360
+ ab_degrade_eb2/01_south_african_002577_clean.wav filter=lfs diff=lfs merge=lfs -text
361
+ ab_degrade_eb2/01_south_african_002577_hard1.wav filter=lfs diff=lfs merge=lfs -text
362
+ ab_degrade_eb2/01_south_african_002577_hard2.wav filter=lfs diff=lfs merge=lfs -text
363
+ ab_degrade_eb2/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_base.wav filter=lfs diff=lfs merge=lfs -text
364
+ ab_degrade_eb2/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_clean.wav filter=lfs diff=lfs merge=lfs -text
365
+ ab_degrade_eb2/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_hard1.wav filter=lfs diff=lfs merge=lfs -text
366
+ ab_degrade_eb2/02_TechinAsia[[:space:]]in[[:space:]]KL,[[:space:]]Sarawak[[:space:]]forging[[:space:]]ahead,_hard2.wav filter=lfs diff=lfs merge=lfs -text
367
+ ab_degrade_eb2/02_emo_neutral_freeform_base.wav filter=lfs diff=lfs merge=lfs -text
368
+ ab_degrade_eb2/02_emo_neutral_freeform_clean.wav filter=lfs diff=lfs merge=lfs -text
369
+ ab_degrade_eb2/02_emo_neutral_freeform_hard1.wav filter=lfs diff=lfs merge=lfs -text
370
+ ab_degrade_eb2/02_emo_neutral_freeform_hard2.wav filter=lfs diff=lfs merge=lfs -text
371
+ ab_degrade_eb2/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_base.wav filter=lfs diff=lfs merge=lfs -text
372
+ ab_degrade_eb2/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_clean.wav filter=lfs diff=lfs merge=lfs -text
373
+ ab_degrade_eb2/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_hard1.wav filter=lfs diff=lfs merge=lfs -text
374
+ ab_degrade_eb2/03_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]Ep.3[[:space:]](Part[[:space:]]2)_hard2.wav filter=lfs diff=lfs merge=lfs -text
375
+ ab_degrade_eb2/03_malayalam_001464_base.wav filter=lfs diff=lfs merge=lfs -text
376
+ ab_degrade_eb2/03_malayalam_001464_clean.wav filter=lfs diff=lfs merge=lfs -text
377
+ ab_degrade_eb2/03_malayalam_001464_hard1.wav filter=lfs diff=lfs merge=lfs -text
378
+ ab_degrade_eb2/03_malayalam_001464_hard2.wav filter=lfs diff=lfs merge=lfs -text
379
+ ab_degrade_eb2/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_base.wav filter=lfs diff=lfs merge=lfs -text
380
+ ab_degrade_eb2/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_clean.wav filter=lfs diff=lfs merge=lfs -text
381
+ ab_degrade_eb2/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_hard1.wav filter=lfs diff=lfs merge=lfs -text
382
+ ab_degrade_eb2/04_EP15[[:space:]]｜[[:space:]]Kenapa[[:space:]]Melayu[[:space:]]Tak[[:space:]]Suka[[:space:]]DAP？,[[:space:]]OKM[[:space:]]_hard2.wav filter=lfs diff=lfs merge=lfs -text
383
+ ab_degrade_eb2/04_south_african_002209_base.wav filter=lfs diff=lfs merge=lfs -text
384
+ ab_degrade_eb2/04_south_african_002209_clean.wav filter=lfs diff=lfs merge=lfs -text
385
+ ab_degrade_eb2/04_south_african_002209_hard1.wav filter=lfs diff=lfs merge=lfs -text
386
+ ab_degrade_eb2/04_south_african_002209_hard2.wav filter=lfs diff=lfs merge=lfs -text
387
+ ab_degrade_eb2/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_base.wav filter=lfs diff=lfs merge=lfs -text
388
+ ab_degrade_eb2/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_clean.wav filter=lfs diff=lfs merge=lfs -text
389
+ ab_degrade_eb2/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_hard1.wav filter=lfs diff=lfs merge=lfs -text
390
+ ab_degrade_eb2/05_\[PART[[:space:]]1\][[:space:]]Telling[[:space:]]My[[:space:]]Children[[:space:]]I[[:space:]]Don't[[:space:]]Hav_hard2.wav filter=lfs diff=lfs merge=lfs -text
391
+ ab_degrade_eb2/05_rainbow_02_slow_base.wav filter=lfs diff=lfs merge=lfs -text
392
+ ab_degrade_eb2/05_rainbow_02_slow_clean.wav filter=lfs diff=lfs merge=lfs -text
393
+ ab_degrade_eb2/05_rainbow_02_slow_hard1.wav filter=lfs diff=lfs merge=lfs -text
394
+ ab_degrade_eb2/05_rainbow_02_slow_hard2.wav filter=lfs diff=lfs merge=lfs -text
395
+ ab_degrade_eb2/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_base.wav filter=lfs diff=lfs merge=lfs -text
396
+ ab_degrade_eb2/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_clean.wav filter=lfs diff=lfs merge=lfs -text
397
+ ab_degrade_eb2/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_hard1.wav filter=lfs diff=lfs merge=lfs -text
398
+ ab_degrade_eb2/06_\#13[[:space:]]Speaking[[:space:]]Up[[:space:]]Against[[:space:]]Sexual[[:space:]]Harassmen_hard2.wav filter=lfs diff=lfs merge=lfs -text
399
+ ab_degrade_eb2/06_uk_ireland_en_011206_base.wav filter=lfs diff=lfs merge=lfs -text
400
+ ab_degrade_eb2/06_uk_ireland_en_011206_clean.wav filter=lfs diff=lfs merge=lfs -text
401
+ ab_degrade_eb2/06_uk_ireland_en_011206_hard1.wav filter=lfs diff=lfs merge=lfs -text
402
+ ab_degrade_eb2/06_uk_ireland_en_011206_hard2.wav filter=lfs diff=lfs merge=lfs -text
403
+ ab_degrade_eb2/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_base.wav filter=lfs diff=lfs merge=lfs -text
404
+ ab_degrade_eb2/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_clean.wav filter=lfs diff=lfs merge=lfs -text
405
+ ab_degrade_eb2/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_hard1.wav filter=lfs diff=lfs merge=lfs -text
406
+ ab_degrade_eb2/07_Seni[[:space:]]bela[[:space:]]diri[[:space:]]wajib[[:space:]]ada[[:space:]]setiap[[:space:]]individu_hard2.wav filter=lfs diff=lfs merge=lfs -text
407
+ ab_degrade_eb2/07_cv_zhcn_001130_base.wav filter=lfs diff=lfs merge=lfs -text
408
+ ab_degrade_eb2/07_cv_zhcn_001130_clean.wav filter=lfs diff=lfs merge=lfs -text
409
+ ab_degrade_eb2/07_cv_zhcn_001130_hard1.wav filter=lfs diff=lfs merge=lfs -text
410
+ ab_degrade_eb2/07_cv_zhcn_001130_hard2.wav filter=lfs diff=lfs merge=lfs -text
411
+ ab_degrade_eb2/08_0001115_base.wav filter=lfs diff=lfs merge=lfs -text
412
+ ab_degrade_eb2/08_0001115_clean.wav filter=lfs diff=lfs merge=lfs -text
413
+ ab_degrade_eb2/08_0001115_hard1.wav filter=lfs diff=lfs merge=lfs -text
414
+ ab_degrade_eb2/08_0001115_hard2.wav filter=lfs diff=lfs merge=lfs -text
415
+ ab_degrade_eb2/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_base.wav filter=lfs diff=lfs merge=lfs -text
416
+ ab_degrade_eb2/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_clean.wav filter=lfs diff=lfs merge=lfs -text
417
+ ab_degrade_eb2/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_hard1.wav filter=lfs diff=lfs merge=lfs -text
418
+ ab_degrade_eb2/08_PODCAST[[:space:]]Borak[[:space:]]Dalam[[:space:]]Galley[[:space:]]EP48[[:space:]]-[[:space:]]Letchu_hard2.wav filter=lfs diff=lfs merge=lfs -text
419
+ ab_degrade_eb2/09_023116_base.wav filter=lfs diff=lfs merge=lfs -text
420
+ ab_degrade_eb2/09_023116_clean.wav filter=lfs diff=lfs merge=lfs -text
421
+ ab_degrade_eb2/09_023116_hard1.wav filter=lfs diff=lfs merge=lfs -text
422
+ ab_degrade_eb2/09_023116_hard2.wav filter=lfs diff=lfs merge=lfs -text
423
+ ab_degrade_eb2/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_base.wav filter=lfs diff=lfs merge=lfs -text
424
+ ab_degrade_eb2/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_clean.wav filter=lfs diff=lfs merge=lfs -text
425
+ ab_degrade_eb2/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_hard1.wav filter=lfs diff=lfs merge=lfs -text
426
+ ab_degrade_eb2/09_Bell[[:space:]]Ngasri[[:space:]]-[[:space:]]OKLETSGO[[:space:]]Video[[:space:]]Podcast[[:space:]](Pa_hard2.wav filter=lfs diff=lfs merge=lfs -text
427
+ ab_degrade_eb2/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_base.wav filter=lfs diff=lfs merge=lfs -text
428
+ ab_degrade_eb2/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_clean.wav filter=lfs diff=lfs merge=lfs -text
429
+ ab_degrade_eb2/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_hard1.wav filter=lfs diff=lfs merge=lfs -text
430
+ ab_degrade_eb2/10_Are[[:space:]]labels[[:space:]]important[[:space:]]in[[:space:]]relationships？[[:space:]]｜_hard2.wav filter=lfs diff=lfs merge=lfs -text
431
+ ab_degrade_eb2/10_tamil_003048_base.wav filter=lfs diff=lfs merge=lfs -text
432
+ ab_degrade_eb2/10_tamil_003048_clean.wav filter=lfs diff=lfs merge=lfs -text
433
+ ab_degrade_eb2/10_tamil_003048_hard1.wav filter=lfs diff=lfs merge=lfs -text
434
+ ab_degrade_eb2/10_tamil_003048_hard2.wav filter=lfs diff=lfs merge=lfs -text
435
+ ab_degrade_eb2/11_rainbow_03_fast_base.wav filter=lfs diff=lfs merge=lfs -text
436
+ ab_degrade_eb2/11_rainbow_03_fast_clean.wav filter=lfs diff=lfs merge=lfs -text
437
+ ab_degrade_eb2/11_rainbow_03_fast_hard1.wav filter=lfs diff=lfs merge=lfs -text
438
+ ab_degrade_eb2/11_rainbow_03_fast_hard2.wav filter=lfs diff=lfs merge=lfs -text
439
+ ab_degrade_eb2/11_uk_ireland_en_008806_base.wav filter=lfs diff=lfs merge=lfs -text
440
+ ab_degrade_eb2/11_uk_ireland_en_008806_clean.wav filter=lfs diff=lfs merge=lfs -text
441
+ ab_degrade_eb2/11_uk_ireland_en_008806_hard1.wav filter=lfs diff=lfs merge=lfs -text
442
+ ab_degrade_eb2/11_uk_ireland_en_008806_hard2.wav filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,142 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-nc-4.0
3
+ tags:
4
+ - speech-restoration
5
+ - speech-enhancement
6
+ - telephony
7
+ - call-center
8
+ - w2v-bert
9
+ - lora
10
+ - sidon
11
+ - dac
12
+ base_model: facebook/w2v-bert-2.0
13
+ pipeline_tag: audio-to-audio
14
+ ---
15
+
16
+ # CallEnhancer — Call-Centre / Telephony Speech Restoration
17
+
18
+ Restore Multi-lingual **narrowband, codec'd, noisy call-centre / telephony speech** (e.g. 8 kHz G.711/GSM phone
19
+ audio) to **clean 48 kHz**.
20
+
21
+ ## Benchmark — CER
22
+
23
+ Character Error Rate (Whisper large-v3, **lower is better**) on a held-out **private call-centre
24
+ test set**. CallEnhancer restores telephony-degraded speech so an ASR reads it far more accurately
25
+ than the raw input or general-purpose speech enhancers.
26
+
27
+ ![CER on a private call-centre test set](assets/callcentre_cer.png)
28
+
29
+ | model | CER % ↓ |
30
+ |---|---:|
31
+ | degraded (original) | 71.48 |
32
+ | voicefixer | 58.27 |
33
+ | resemble-enhance | 54.96 |
34
+ | sidon-v0.1 | 41.06 |
35
+ | **CallEnhancer-small** (this repo, open source) | **38.91** |
36
+ | **CallEnhancer-base** | **19.01** |
37
+
38
+ This repository hosts **CallEnhancer-small** (open source); **CallEnhancer-base** is the larger model.
39
+
40
+ ## Quick start — infer from the HF checkpoint
41
+
42
+ ```bash
43
+ pip install torch torchaudio "transformers>=4.56" "descript-audio-codec>=1.0.0" soundfile "huggingface_hub[cli]"
44
+
45
+ # pull the CLI + the two slim checkpoints from the Hub
46
+ hf auth login # private repo: log in first (or export HF_TOKEN=hf_...)
47
+ hf download Scicom-intl/CallEnhancer \
48
+ infer_callcentre.py fe_callcentre/fe_adapter_full.pt \
49
+ decoder_callcentre/decoder_only.pt --local-dir CallEnhancer
50
+
51
+ cd CallEnhancer && python infer_callcentre.py \
52
+ --input your_call.wav --out-dir out \
53
+ --fe-adapter fe_callcentre/fe_adapter_full.pt \
54
+ --decoder decoder_callcentre/decoder_only.pt \
55
+ --chunk 0 --device cuda # --chunk 0 = NO chunking (default single pass); --device cpu if no GPU
56
+ # -> out/your_call_restored48k.wav (clean 48 kHz) + out/your_call_orig48k.wav (A/B)
57
+ ```
58
+
59
+ Prefer Python (load weights from the Hub with `hf_hub_download`)? See **[Python](#python-pull-weights-from-the-hub)** below.
60
+
61
+ > **Status:** CallEnhancer-small is fully trained and restores real 8 kHz call-centre audio well.
62
+
63
+ ## Files
64
+
65
+ Use the **current-run** checkpoints under `fe_callcentre/` and `decoder_callcentre/`:
66
+
67
+ | path | role | size |
68
+ |---|---|---|
69
+ | `fe_callcentre/fe_adapter_full.pt` | **FE adapter (inference)** — 144 tensors: 96 LoRA + 48 trained `output_dense` biases | ~63 MB |
70
+ | `decoder_callcentre/decoder_only.pt` | **decoder (inference)** — 188M DAC decoder | ~0.75 GB |
71
+ | `fe_callcentre/last.pt`, `decoder_callcentre/last.pt` | raw checkpoints (resume training) | ~2.5 / 2.8 GB |
72
+ | `infer_callcentre.py` | inference CLI (below) | — |
73
+
74
+ For inference you only need the two slim files + `infer_callcentre.py`. *(Root-level
75
+ `fe_adapter_full.pt` / `decoder_only.pt` are from an earlier run and are superseded.)*
76
+
77
+ ## End-to-end example (straight from HuggingFace)
78
+
79
+ ```bash
80
+ pip install torch torchaudio "transformers>=4.56" "descript-audio-codec>=1.0.0" soundfile "huggingface_hub[cli]"
81
+
82
+ # pull the CLI + the two slim checkpoints, straight from this repo
83
+ hf auth login # private repo: log in first (or export HF_TOKEN=hf_...)
84
+ hf download Scicom-intl/CallEnhancer \
85
+ infer_callcentre.py \
86
+ fe_callcentre/fe_adapter_full.pt \
87
+ decoder_callcentre/decoder_only.pt \
88
+ --local-dir CallEnhancer
89
+ cd CallEnhancer
90
+
91
+ # restore your audio end-to-end
92
+ python infer_callcentre.py \
93
+ --input your_call.wav \
94
+ --out-dir out \
95
+ --fe-adapter fe_callcentre/fe_adapter_full.pt \
96
+ --decoder decoder_callcentre/decoder_only.pt \
97
+ --chunk 0 --device cuda # --chunk 0 = NO chunking (single straight pass, default); --device cpu if no GPU
98
+ ```
99
+
100
+ Outputs:
101
+ - `out/your_call_restored48k.wav` — the restored **clean 48 kHz** speech.
102
+ - `out/your_call_orig48k.wav` — the input, naively upsampled to 48 kHz (no model), for an A/B listen.
103
+
104
+ `--input` accepts a **file or a directory** (`.wav/.flac/.mp3/.ogg/.opus/.m4a`). Stereo (e.g.
105
+ agent/customer on separate channels) is restored per channel and recombined.
106
+
107
+ ### Python (pull weights from the Hub)
108
+
109
+ ```python
110
+ import numpy as np, soundfile as sf, torch, torchaudio
111
+ from huggingface_hub import hf_hub_download
112
+ from transformers import AutoFeatureExtractor, Wav2Vec2BertModel
113
+ import dac
114
+
115
+ REPO, SSL, FE_SR, SR_OUT = "Scicom-intl/CallEnhancer", "facebook/w2v-bert-2.0", 16000, 48000
116
+ dev = torch.device("cuda" if torch.cuda.is_available() else "cpu")
117
+
118
+ ck = torch.load(hf_hub_download(REPO, "fe_callcentre/fe_adapter_full.pt"), map_location="cpu")
119
+ ad, scale = ck["adapter"], ck["lora_alpha"] / ck["r"]
120
+ fe = Wav2Vec2BertModel.from_pretrained(SSL, num_hidden_layers=ck.get("layers", 24), layerdrop=0.0)
121
+ sd = fe.state_dict() # merge LoRA -> base (no peft needed)
122
+ for p in sorted({k[:-len(".lora_A.default.weight")] for k in ad if k.endswith(".lora_A.default.weight")}):
123
+ sd[p+".weight"] = sd[p+".weight"].float() + scale * (ad[p+".lora_B.default.weight"].float() @ ad[p+".lora_A.default.weight"].float())
124
+ if p+".base_layer.bias" in ad: sd[p+".bias"] = ad[p+".base_layer.bias"].to(sd[p+".bias"].dtype)
125
+ fe.load_state_dict(sd); fe.to(dev).eval()
126
+
127
+ dck = torch.load(hf_hub_download(REPO, "decoder_callcentre/decoder_only.pt"), map_location="cpu")
128
+ dec = dac.model.dac.Decoder(input_channel=1024, channels=dck.get("dec_channels", 3072), rates=[8,5,4,3,2])
129
+ dec.load_state_dict(dck["decoder"]); dec.to(dev).eval()
130
+ proc = AutoFeatureExtractor.from_pretrained(SSL)
131
+
132
+ @torch.no_grad()
133
+ def restore(path, out="restored48k.wav"): # single straight pass
134
+ x, sr = sf.read(path, dtype="float32"); x = x.mean(1) if x.ndim > 1 else x
135
+ if sr != FE_SR: x = torchaudio.functional.resample(torch.from_numpy(x)[None], sr, FE_SR)[0].numpy()
136
+ x = x / (np.abs(x).max() + 1e-9) * 0.95
137
+ feats = {k: v.to(dev) for k, v in proc(x, sampling_rate=FE_SR, return_tensors="pt").items()}
138
+ y = dec(fe(**feats).last_hidden_state.transpose(1, 2)).squeeze().float().cpu().numpy()
139
+ sf.write(out, y / (np.abs(y).max() + 1e-9) * 0.97, SR_OUT); print("wrote", out)
140
+
141
+ restore("your_call.wav") # <-- your own telephony/call-centre audio
142
+ ```
assets/callcentre_cer.png ADDED

Git LFS Details

  • SHA256: 857dc26bb4b97b90ad2d6d803e80db400e30b5e76ade82fcc301f3d7bf83c7af
  • Pointer size: 131 Bytes
  • Size of remote file: 101 kB
callhome_benchmark/callhome_benchmark.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11b27646d8b9127f9ca5cfade46b38d70e289dac855c177ad1bce6c2764f1ad2
3
+ size 10681510070
callhome_benchmark/results.md ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # CallHome restoration benchmark (UTMOS)
2
+
3
+ talkbank/callhome, 30 s chunks, <=100/language. UTMOS via faster-UTMOSv2 (higher = better). No transcripts in this dataset, so no CER — perceptual quality only. Δ is vs the untouched **degraded** (original) audio.
4
+
5
+ | model | UTMOS (↑) | Δ vs orig | deu | eng | jpn | spa | zho | n |
6
+ |---|---:|---:|---:|---:|---:|---:|---:|---:|
7
+ | degraded *(original)* | 1.538 | — | 1.541 | 1.814 | 1.545 | 1.221 | 1.570 | 500 |
8
+ | ours | 2.195 | +0.657 | 2.179 | 2.316 | 2.260 | 1.847 | 2.374 | 500 |
9
+ | sidon_v01 | 2.711 | +1.173 | 2.720 | 3.021 | 2.803 | 2.205 | 2.806 | 500 |
10
+ | voicefixer | 1.891 | +0.353 | 1.765 | 1.989 | 1.824 | 1.900 | 1.976 | 500 |
11
+ | resemble_enhance | 2.311 | +0.774 | 2.378 | 2.439 | 2.345 | 2.003 | 2.392 | 500 |
decoder_callcentre/decoder_only.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5af53b93e918b279aa1a34179f164e0cecda6b532f2b338756b31c1e0d6e9c1
3
+ size 750806394
decoder_callcentre/last.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20314539fd278e06672473158c06117f8327981cb44bdece0a78f207fc85b2cc
3
+ size 2763003541
decoder_only.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd531628973fe2b97a80340954c8789655bea03e219b81f71c545e0ec3278c68
3
+ size 750805782
decoder_train_state.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f2e9a7e09f486b54ec6212ca24f98f69eef9a7977033f0e8af0854e67f26e71
3
+ size 2763003541
examples/callhome_deu.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7a1e7ba05cfa60f5771046b7d87be367594c3d337923ed5c3a8f2a3f342567e
3
+ size 180302
examples/callhome_deu_restored48k_step108000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d01e497e9558675ea876f0774fb05a8b60195e67e3b57e2a296df903cc2be934
3
+ size 1081388
examples/callhome_deu_restored48k_step120000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:827eba9f4cef42c88ffcc4b0cf30afd153c0c2f7787afddea1b63740aadf3ebc
3
+ size 1081388
examples/callhome_deu_restored48k_step132000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00e166efff4fd0090f9fdf22400e6a69ed44ac56587e55af185819297e03f4d9
3
+ size 1081388
examples/callhome_deu_restored48k_step144000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b6f50d59574a32d9471ac1f525da0ae3d7a9e178cb488e75b6e41baef8e5267
3
+ size 1081388
examples/callhome_deu_restored48k_step148000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c56566ddd9a4138465ac9eb49d469bffeac3efcf1756c09a9d19adb13948f0f3
3
+ size 1081388
examples/callhome_deu_restored48k_step158000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c16677c257780491410d88cb3d364ec12b05f29eaa5d46acb1ad95d1bc6d00a
3
+ size 1081388
examples/callhome_deu_restored48k_step168000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd0f97a81ef0ab70c7e7a38025030bf19fe9f5292d0ee7e5b89e9d1c426fba8e
3
+ size 1081388
examples/callhome_deu_restored48k_step180000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fda61576a961bbd39d59ef6c475763636c51cdc5eafc684f718d82aa5218d76d
3
+ size 1081388
examples/callhome_deu_restored48k_step194000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4540f5457626397fe5ee8c2d90815cf4dfa332a7011742152fc2359dd5e6168
3
+ size 1081388
examples/callhome_deu_restored48k_step212000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ead7b065a07ad479ef3955fe3d9eae9c0da3cb2e8f5ab9e6bc95d79f6df3545
3
+ size 1081388
examples/callhome_deu_restored48k_step226000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b46087b81a7d594001b78dd8433a5d3c1e0db60b04ed21ad6f5a15dffd0bd185
3
+ size 1081388
examples/callhome_deu_restored48k_step242000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e553cbe038dba828b1e87e597ab49371fef9ac34866bfd601f873291bf9a0aaa
3
+ size 1081388
examples/callhome_deu_restored48k_step262000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd6d6786577b3329319f1c81cef689464dd5f320fe8960b994a01d7cef7b0eba
3
+ size 1081388
examples/callhome_deu_restored48k_step286000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2d04b36645c027c9f0bcee14242100f4f757234bf716a791a458f6f193a415c
3
+ size 1081388
examples/callhome_deu_restored48k_step310000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b1f22156fa459d03f23bab10995bc77811560049bd43f2060420fac45d33291
3
+ size 1081388
examples/callhome_deu_restored48k_step332000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6fdd0ed28b3beb0495f0e6c491106db9a302085cdac9572f390510263635d442
3
+ size 1081388
examples/callhome_deu_restored48k_step342000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d908e0a83cba995696d0d72c93bb6e4031e0006c51826cb0ea807a87cfb2fad
3
+ size 1081388
examples/callhome_deu_restored48k_step356000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbe7f2b8f1c6d5edb37a13b563923ff59545285533f0ab97b45d33cd7a65e6ca
3
+ size 1081388
examples/callhome_deu_restored48k_step382000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:445ee79095b9fca745255036d168eee9939494b709c7189005ab1db57434e4b3
3
+ size 1081388
examples/callhome_deu_restored48k_step392000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0157f56a25ced1801fddfdec11867c21e0206989cde203610f678a9631375f2
3
+ size 1081388
examples/callhome_deu_restored48k_step408000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33dec41e9c8bb540c0df69450cc87a022a57d0c668bd466f3f4d4c8afa6f8143
3
+ size 1081388
examples/callhome_deu_restored48k_step432000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:585a210e2200dd1a4cdf2f5a6b4f2415e6964b9ed46ad8ba13486ff2bbff5b19
3
+ size 1081388
examples/callhome_deu_restored48k_step444000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8abcd21ea9ddc4be61a77f0cc84065e32e8eb6a8d63541144a85f73b156c7d0f
3
+ size 1081388
examples/callhome_deu_restored48k_step458000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8965c2f082c0218425459f9a31e01c6f69db5aa87fac28f1dfc6d890f60fda0f
3
+ size 1081388
examples/callhome_deu_restored48k_step46000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8aa1d7280f17ca2b278d7c2ddb782f9b318aca9edc09812df65cacb80f875382
3
+ size 1081388
examples/callhome_deu_restored48k_step484000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6823d46be89798f2df6839a9d6cb9dc6834ecdf4f16f0048c586697f8301a4b
3
+ size 1081388
examples/callhome_deu_restored48k_step492000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dab5433a9017381daf128354b54fc0e936cb9b51e7190da9e0a2dec13df0194f
3
+ size 1081388
examples/callhome_deu_restored48k_step500000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39a612e2c37c2203c7b633e4449cb8daccfd5552012c76517e815afecc7ff983
3
+ size 1081388
examples/callhome_deu_restored48k_step56000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62d9aca88bafe946d4cf48227482f5c61c6fa37b7d0da89dd417dfdd3d64c4bc
3
+ size 1081388
examples/callhome_deu_restored48k_step66000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63ce78c79cb91a28848f9d5824c611be478884c4b3247a5cd0b20225fffff1da
3
+ size 1081388
examples/callhome_deu_restored48k_step78000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c37c4c508590b7cf5f764c262b1326c1c7743e44676d7f42af6808da4e8dd05
3
+ size 1081388
examples/callhome_deu_restored48k_step94000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c68d1412469a1584da8613ff559cf218f21ac08ad7c1a7b629b9ce9edb8ed30
3
+ size 1081388
examples/callhome_eng.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a5733271526d5a22d7f2aed68fe1d7880d0a65c81e4c3b6da0888da5c2cb4b1
3
+ size 131150
examples/callhome_eng_restored48k_step108000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:acdd0dd22408512e61358482a303df9dff55212e31d75de87a0b61ca43ff3c88
3
+ size 786476
examples/callhome_eng_restored48k_step120000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc954cde6da44623966a7924309e71249298091802960866fd0212822630f8bd
3
+ size 786476
examples/callhome_eng_restored48k_step132000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c874799c57a1641c7ca881a37afa01e722edaa404f79331eaa0c74aa62ea546c
3
+ size 786476
examples/callhome_eng_restored48k_step144000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c448230b3db5d1bb142a02edf54b2f66423ec28850a28394e03e22860567a3f
3
+ size 786476
examples/callhome_eng_restored48k_step148000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45b8069f9e51b40003903b74f8e2e2d91612d167835e877fc906637229596ac8
3
+ size 786476
examples/callhome_eng_restored48k_step158000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b50adf2b2fa1537e575aa9e25012ec8ae221c8629e35446b27beb58bd968208
3
+ size 786476
examples/callhome_eng_restored48k_step168000.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1c38a98e938a3de64cc07c7a8e116fe26bbb5313c3590f07d16fae68c52100a
3
+ size 786476