Fast-Forward Video Based on Semantic Extraction
This addresses the challenge of boring egocentric videos for users of smartphones and wearable devices, though it is incremental as it builds on existing fast-forward techniques.
The authors tackled the problem of making long, unedited egocentric videos more appealing by proposing a method to create fast-forward videos based on semantic frame selection, resulting in videos that outperform state-of-the-art in semantic information and are more pleasant to watch.
Thanks to the low operational cost and large storage capacity of smartphones and wearable devices, people are recording many hours of daily activities, sport actions and home videos. These videos, also known as egocentric videos, are generally long-running streams with unedited content, which make them boring and visually unpalatable, bringing up the challenge to make egocentric videos more appealing. In this work we propose a novel methodology to compose the new fast-forward video by selecting frames based on semantic information extracted from images. The experiments show that our approach outperforms the state-of-the-art as far as semantic information is concerned and that it is also able to produce videos that are more pleasant to be watched.