Repository navigation
VoiceFlow 0.3.11
·
194 commits
to dfde0e2bce61f6627acf3e0ef2da4f188245bf6f
since this release
VoiceFlow 0.3.11
中文
- 20 秒以内的语音可在说话停顿时提前识别;如果停下前没有继续说话,直接使用后台结果,减少等待。
- 改进长语音分段衔接,减少接缝处的多余重复、误删真实重复词和中文多余空格。
- 静音听写可正常取消;录音胶囊的声波位置固定,文字在右侧居中,展开更平滑。
- 更新 anyio、certifi、filelock 和 typing_extensions 依赖。
- 官网增加页脚链接,并在从作者个人网站进入时显示返回入口;这些改动仅影响官网。
English
- For recordings under 20 seconds, recognition may run during a pause and reuse the result when no further speech follows.
- Improve long-recording joins to reduce duplicate text, preserve intentional repetitions, and avoid extra spaces at Chinese seams.
- Cancel silent dictation cleanly; the capsule keeps its waveform anchored while text stays centered in the right-hand area as it expands.
- Update anyio, certifi, filelock, and typing_extensions.
- Add a website footer link and a return link for visitors arriving from the author's personal site; these changes affect the website only.
Windows installer is unsigned.