tyiannak / multimodal_movie_analysis

A Python Library for Multimodal Analysis of Movies and Content-based Movie Recommendation
25 stars 8 forks source link

Shot Evaluation: select 20 videos from different movies #11

Closed tyiannak closed 3 years ago

tyiannak commented 3 years ago

@kbogas note: try to focus on videos from directors / styles that have frequent shots + 2-3 long shots

tyiannak commented 3 years ago

Deliverable: 20 videos in gdrive

tyiannak commented 3 years ago

@kbogas comment: try to cover all available types of shots in this small dataset so that we know that accuracy is more-or-less acceptable in all cases. @apoman38 pls paste your current list of classes here

apoman38 commented 3 years ago

Link for the file with shot types: https://docs.google.com/presentation/d/1fnnMXfxwrQHA7LolTK4_6ptf-Lptrz5-4_mas6JJqvQ/edit?usp=sharing

tyiannak commented 3 years ago

Data is available here (by @electrasif ) https://drive.google.com/drive/folders/13IY9DeJoGruDSqftcarVt2SsZjyQDFHW

@apoman38 please review

apoman38 commented 3 years ago

Good job @electrasif ! You made a good choice of videos and I think that is enough for us to work on at the moment.

tyiannak commented 3 years ago

Cool, i am closing that. And we can continue in the annotation itself right away. @electrasif @apoman38 i will assign you videos to annotate for shot boundaries in the other task