Repository for project at Machine learning for NLP course at Chalmers University of Technology.
Check out the project report for description of the entire project: Project report
We present a comprehensive investigation into the problem of automatically generating timestamps for videos. To address this challenge, we propose and evaluate two different automatic pipelines for timestamp generation. Our first approach combines automatic speech recognition with state-of-the-art language models to generate suggestions for timestamps. The results of our study demonstrate that our pipeline generates adequate keywords for the video but the timing of the times- tamps remains a challenge. Overall, our study offers a promising approach for automatically generating timestamps for videos, and in particular it highlights the potential for further work in this area. Our code for the project is available at: github.com/SannaPersson/Automatic-Youtube-Timestamps