4 ms·
Why did a search engine for movies need the actual movie files, and not just the metadata? Or was it actually making the movies themselves available?
by Rotten194 8y ago
Why did a search engine for movies need the actual movie files, and not just the metadata? Or was it actually making the movies themselves available?
- techdragon 8y agoI would guess it’s in order to build the hash/IDs/searchable dataset that can be used to match the movies. Think Shazam style fingerprints but for movies. I was looking into this sort of thing recently for a project. Very interesting technology space with some genuine direct benefits coming out of the application of machine learning.
- NoblePublius 8y agoCorrect
- haser_au 8y agoA guess would be more movie meta data than you can think of; - show me movies with Swiss Alps settings - show me movies with the quote "I'll be back" - tell me which part of Back to the Future shows me where they have hoverboards - actor/director/artist/crew: for my IMDB profile, tell me when my name is on screen (start credits, end credits)
- chrismeller 8y agoI'm guessing something along the lines of Morbotron [1]. Not the actual use case for the screens/subs/etc. but more the data required to do either of them: frame by frame images for analysis with timestamps and subtitles. 1: https://morbotron.com/ https://morbotron.com/
- reaperducer 8y agoWhy did a search engine for movies need the actual movie files, and not just the metadata? Someone's got to build that metadata. Especially back in 2009, you couldn't just download it from some public database. Databases have to be built.
- NoblePublius 8y agoWe needed high quality video files to create the metadata. Machine vision wasn’t good enough yet so we did it the old fashion way. Interns and lots of typing in real time