Master thesis : Content-aware retargeting of broadcast videos
Pagliarello, Lorenzo
Promoteur(s) :
Geurts, Pierre
;
Barnich, Olivier
Date de soutenance : 27-jui-2022/28-jui-2022 • URL permanente : http://hdl.handle.net/2268.2/14332
Détails
Titre : | Master thesis : Content-aware retargeting of broadcast videos |
Titre traduit : | [fr] Reciblage intelligent du format de vidéos télédiffusées |
Auteur : | Pagliarello, Lorenzo ![]() |
Date de soutenance : | 27-jui-2022/28-jui-2022 |
Promoteur(s) : | Geurts, Pierre ![]() Barnich, Olivier |
Membre(s) du jury : | Van Droogenbroeck, Marc ![]() Wehenkel, Louis ![]() |
Langue : | Anglais |
Nombre de pages : | 96 |
Mots-clés : | [en] Video retargeting [en] Video saliency detection [en] One dimensional cropping [en] Dynamic programming |
Discipline(s) : | Ingénierie, informatique & technologie > Sciences informatiques |
Institution(s) : | Université de Liège, Liège, Belgique |
Diplôme : | Master en ingénieur civil en informatique, à finalité spécialisée en "intelligent systems" |
Faculté : | Mémoires de la Faculté des Sciences appliquées |
Résumé
[en] Video retargeting, or the challenge of transforming a video from one aspect ratio to another, has become a source of great interest in recent years. While the de-facto standard for filming productions has been 16:9 for a long time, the growth of social media and the broadening of screen sizes demand for an automatic conversion procedure. With this thesis, we provide an overview of the current practices for this field both in the literature and in the industry. We discuss why one dimensional cropping should be preferred over other hybrid techniques in the context of the broadcast industry.
Resulting from this study, we introduce our own modular framework composed of two subsequent computational blocs. On one hand, the first module comprises a state-of-the-art video saliency detection model which locates and quantifies relevant information. As part of our contributions, we build our own saliency dataset called EVS-Sal and fine-tune the deep network to specialize its detections for soccer content. On the other hand, the second module is responsible for the selection of cropped salient information while ensuring temporal consistency. For this purpose, we explore both global and local optimizations respectively with the dynamic programming paradigm and with a “select and filter” approach.
Finally, we show that our methods outperform current one dimensional retargeting algorithms on a variety of general videos. Additionally, we extend this analysis with the creation of our own soccer retargeting dataset called EVS-Ret. With the latter, we demonstrate that our framework brings results near inter-human agreement and that the semantics of soccer are correctly captured by the re-trained saliency model.
Fichier(s)
Document(s)
![File](/static/img/item/file.png)
![Accès privé Access](/static/img/item/file/pdf_priv.png)
Description: PDF version of the master thesis.
Taille: 42.58 MB
Format: Adobe PDF
![File](/static/img/item/file.png)
![Accès privé Access](/static/img/item/file/pdf_priv.png)
Description: One page PDF summary of the master thesis.
Taille: 231.3 kB
Format: Adobe PDF
Annexe(s)
![File](/static/img/item/file.png)
![Accès privé Access](/static/img/item/file/pdf_priv.png)
Description: Source code. Large files (data, model weights, ...) are omitted and stored in EVS facilities.
Taille: 67.96 MB
Format: Unknown
Citer ce mémoire
L'Université de Liège ne garantit pas la qualité scientifique de ces travaux d'étudiants ni l'exactitude de l'ensemble des informations qu'ils contiennent.