The LectureCut Generator is a module of the LectureCut project. It is responsible for analyzing the lecture video to find parts where no one is speaking. From these parts, a cut list is generated, which can be used by the render module to cut the lecture video into its final form.
Internally the Generator uses avcodec and avformat from the FFmpeg project to extract the audio from the video and libfvad to detect speech in the audio. The Generator is written in C++ to allow fast and efficient processing of videos.
The LectureCut Generator is licensed under the MIT License.