BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Aesthetics-oriented video generation and editing
DTSTART:20220823T100000
DTEND:20220823T120000
DTSTAMP:20260408T115523Z
UID:f64f4e826c1defa2908408e7f882063e2bbfa52f722f98b420a0fc5b
CATEGORIES:Conferences - Seminars
DESCRIPTION:Martin Nicolas EVERAERT\nEDIC candidacy exam\nExam president: 
 Prof. Tanja Käser\nThesis advisor: Prof. Sabine Süsstrunk\nThesis co-adv
 isor: Dr Radhakrishna Achanta\nCo-examiner: Dr Mathieu Salzmann\n\nAbstrac
 t\nThe average time spent watching online videos increases every year\, ac
 ross all demographics. Videos are more engaging and are shared twice as mu
 ch as other types of media. However\, making or editing such videos can be
  expensive and time-consuming. Our research goal is to propose solutions b
 ased on machine learning and computational aesthetics to automate steps in
  the creation and editing of videos that are appealing and of interest for
  the viewer.\nIn this research proposal\, we discuss three existing works 
 and how they relate to our research. We first examine how generative adver
 sarial networks (GANs) can be used to generate videos and what are their l
 imitations.Then\, we take a look at an example of data collection and anno
 tation process\, allowing training of models for video aesthetics and mess
 age understanding.\nFinally\, we discuss a framework to navigate GANs' lat
 ent space to improve aesthetics.\n\nBackground papers\n[1] Temporal Shift
  GAN for Large Scale Video Generation\, Munoz et al.\, WACV 2021\, https:
 //arxiv.org/abs/2004.01823\n[2] GANalyze: Toward Visual Definitions of Co
 gnitive Image Properties\, Goetschalckx et al.\, ICCV 2019\, https://ar
 xiv.org/abs/1906.10112\n[3] Automatic Understanding of Image and Video Adv
 ertisements\, Hussain et al.\, CVPR 2017\, https://arxiv.org/abs/1707.03
 067\n 
LOCATION:BC 329 https://plan.epfl.ch/?room==BC%20329
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR