TY - DATA T1 - Data and code underlying the paper: "Can we predict the Most Replayed data of video streaming platforms?" PY - 2024/05/24 AU - Alessandro Duico AU - Ombretta Strafforello AU - Jan van Gemert UR - DO - 10.4121/0ca18691-3fef-4c9c-9080-12b20daae62a.v1 KW - Most Replayed Data KW - Video streaming KW - Video understanding KW - Computer vision N2 -

Predicting which specific parts of a video users will replay is important for several applications, including targeted advertisement placement on video platforms and assisting video creators. In this work, we explore whether it is possible to predict the Most Replayed (MR) data from YouTube videos. To this end, we curate a large video benchmark, the YTMR500 dataset, which comprises 500 YouTube videos with MR data annotations. We evaluate Deep Learning (DL) models of varying complexity on our dataset and perform an extensive ablation study. In addition, we conduct a user study to estimate the human performance on MR data prediction. Our results show that, although by a narrow margin, all the evaluated DL models outperform random predictions. Additionally, they exceed human-level accuracy. This suggests that predicting the MR data is a difficult task that can be enhanced through the assistance of DL. In this repository, we provide our code and dataset. The code includes our trained and tested models, our user studies and results analysis. The YTMR500 dataset is provided through an H5 file.

ER -