Skip to content

FunASR-powered video transcription, subtitle generation, and LLM-assisted clipping tool with a local Gradio UI.

update audio recog

shixian.shi682e97ffb25d
diff --git a/paraclipper/launch.py b/paraclipper/launch.py
index db77b33..bb60299 100644
--- a/paraclipper/launch.py
+++ b/paraclipper/launch.py
@@ -15,9 +15,10 @@ if __name__ == "__main__":
)
audio_clipper = VideoClipper(funasr_model)
- def audio_recog(audio_input, sd_switch):
+ def audio_recog(audio_input, sd_switch, hotwords):
+ # import pdb; pdb.set_trace()
print(audio_input)
- return audio_clipper.recog(audio_input, sd_switch)
+ return audio_clipper.recog(audio_input, sd_switch, hotwords=hotwords)
def audio_clip(dest_text, audio_spk_input, start_ost, end_ost, state):
return audio_clipper.clip(dest_text, start_ost, end_ost, state, dest_spk=audio_spk_input)
@@ -83,8 +84,9 @@ if __name__ == "__main__":
gr.Examples(['https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/%E8%AE%BF%E8%B0%88.mp4'],
[video_input],
label='说话人切分示例 SD Demo')
- video_sd_switch = gr.Radio(["no", "yes"], label="👥是否区分说话人 Recognize Speakers", value='no')
- hotwords_input = gr.Textbox(label="🚒热词 Hotwords")
+ with gr.Row():
+ video_sd_switch = gr.Radio(["no", "yes"], label="👥是否区分说话人 Recognize Speakers", value='no')
+ hotwords_input = gr.Textbox(label="🚒热词 Hotwords")
recog_button2 = gr.Button("👂识别 Recognize")
video_text_output = gr.Textbox(label="✏️识别结果 Recognition Result")
video_srt_output = gr.Textbox(label="📖SRT字幕内容 RST Subtitles")
@@ -110,7 +112,9 @@ if __name__ == "__main__":
with gr.Column():
audio_input = gr.Audio(label="🔊音频输入 Audio Input")
gr.Examples(['https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/%E9%B2%81%E8%82%83%E9%87%87%E8%AE%BF%E7%89%87%E6%AE%B51.wav'], [audio_input])
- audio_sd_switch = gr.Radio(["no", "yes"], label="👥是否区分说话人 Recognize Speakers", value='no')
+ with gr.Row():
+ audio_sd_switch = gr.Radio(["no", "yes"], label="👥是否区分说话人 Recognize Speakers", value='no')
+ hotwords_input2 = gr.Textbox(label="🚒热词 Hotwords")
recog_button1 = gr.Button("👂识别 Recognize")
audio_text_output = gr.Textbox(label="✏️识别结果 Recognition Result")
audio_srt_output = gr.Textbox(label="📖SRT字幕内容 RST Subtitles")
@@ -127,7 +131,7 @@ if __name__ == "__main__":
audio_srt_clip_output = gr.Textbox(label="📖裁剪部分SRT字幕内容 Clipped RST Subtitles")
recog_button1.click(audio_recog,
- inputs=[audio_input, audio_sd_switch],
+ inputs=[audio_input, audio_sd_switch, hotwords_input2],
outputs=[audio_text_output, audio_srt_output, audio_state])
clip_button1.click(audio_clip,
inputs=[audio_text_input, audio_spk_input, audio_start_ost, audio_end_ost, audio_state],
diff --git a/paraclipper/videoclipper.py b/paraclipper/videoclipper.py
index da412c5..301cf56 100644
--- a/paraclipper/videoclipper.py
+++ b/paraclipper/videoclipper.py
@@ -69,9 +69,15 @@ class VideoClipper():
offset_b, offset_e = 0, 0
log_append = "(Bracket detected in dest_text but offset time matching failed)"
_dest_text = _dest_text[:_dest_text.find('[')]
+ else:
+ log_append = ""
+ offset_b, offset_e = 0, 0
_dest_text = pre_proc(_dest_text)
ts = proc(recog_res_raw, timestamp, _dest_text)
- for _ts in ts: all_ts.append(_ts)
+ for _ts in ts: all_ts.append([_ts[0]+offset_b*16, _ts[1]+offset_e*16])
+ if len(ts) > 1 and match:
+ log_append += '(offsets detected but No.{} sub-sentence matched to {} periods in audio, \
+ offsets are applied to all periods)'
else:
for _dest_spk in dest_spk.split('#'):
ts = proc_spk(_dest_spk, state['sd_sentences'])