A Self-Adapting Approach to Topic Boundary Detection in Video Subtitles
Xin Qi, Shujin Lin, Baoquan Zhao, Xiaonan Luo · 2016
Our goal is to automatically partitioning video subtitles into coherent segments which reflect subtopic structure of the video subtitles. Our approach enlists self-adapting text segmentation to adapt both long and short video subtitles. We evaluate our algorithm by comparing the automatically detected boundaries with the ground truth results by human judgements, and computing precision.