About the Assistant

'Video Translation Assistant' is an end-to-end video localization tool. It supports video upload → optional watermark/subtitle area selection → audio extraction → speech / text recognition → AI translation → subtitle review → text-to-speech (TTS) → final video rendering.

The system uses a two-stage workflow so that you can review and adjust translations and voices before final rendering, ensuring high-quality localized videos.

Standard Workflow
1Upload & Create Task

Click 'New Task', enter a task name, and choose the target language (English, Japanese, etc.).
Tip: If the original video has hard subtitles, we recommend using OCR mode. If there are no subtitles, please choose ASR mode.

2Preprocessing (optional)

After the upload succeeds, the status becomes 'Uploaded'. If you need to remove a watermark or specify the subtitle recognition area, click Edit watermark or subtitle area in the list to draw the region.

3Start Stage 1 (translation)

Select the task and click 'Start Stage 1'. The system will automatically run speech / text recognition and translation.
During processing the status is Processing, and after completion it becomes To review.

4Manual review & editing

When the task is in the 'To review' state, click 'Edit subtitle settings / review'.
On this page you can fix mistranslations, adjust timing, choose TTS voices, and configure subtitle styles (color, size, etc.).

5Start Stage 2 (rendering)

After confirming that subtitles are correct, click 'Start Stage 2' in the task list.
The system will generate dubbing according to your settings and render the final video. When finished, the status becomes Success.

FAQ

Insert \n where you want a line break. For example: First line\nSecond line. In the rendered video this will appear as two lines.

If the status shows Failed, you can try clicking Restart in the toolbar. If the problem persists, please contact the administrator to check logs.

  • Re-render (Refinalize): Available only when the task is in the 'Success' state. If you are not satisfied with subtitle styling or want to tweak translations, edit them and click this button. The system will skip recognition and quickly regenerate the video.
  • Restart: Clears all generated subtitles and intermediate files and runs the whole workflow again from Stage 1 (recognition and translation).

When the task status is Success, select the task and click the green Download Video button.
Status Legend
  • Uploaded Video has been stored and is waiting for Stage 1.
  • Queued Task has been submitted and is waiting for backend resources.
  • Processing Recognition, translation or rendering is in progress.
  • To review Stage 1 finished; subtitles are waiting for manual review.
  • Success All steps finished; the video can be downloaded.
  • Failed An error occurred. Please retry or contact the administrator.
Button Quick Reference
Button Description
Start Stage 1 Start speech / text recognition and translation.
Edit subtitle settings / review Manually review subtitles and adjust styles and voices.
Edit watermark / subtitle area Configure OCR detection or watermark removal regions.
Start Stage 2 After confirmation, start final rendering.
Re-render Only regenerate the final video (skip recognition).
Stop Force stop a task that is currently running.
Feedback & Discussion
WeChat: fgai66 (please note 'Translation Assistant')
Email: ai_chatbot@163.com
QQ Group: 1061043049
Douyin: Ye Laoshi on AI