Valossa Video Search Uses AI To Let Users Search For Movies When They Forget Titles

Meet Val.ai: an advanced video analysis platform that helps you identify films based on themes and concepts.
Meet Val.ai: an advanced video analysis platform that helps you identify films based on themes and concepts. Valossa

The struggle is real when you’re thinking of a movie and can remember every detail (the plot, the actors, the director) except for the title. The struggle gets worse if you can only recollect tidbits and not any specific cast members, as even IMDB will not be able to lend a helping hand.

In such dire situations, a website called Valossa Video Search can help using artificial intelligence. Created by a Finland-based tech firm, the technology is in its beta phase and has been created based on research conducted at the University of Oulu. The AI system has the ability to analyze descriptions and offer up a title based on concepts, places, themes and actors.

“Valossa AI is capable of analyzing any video stream automatically identifying thousands of concepts and themes,” says the website. “The AI based deep video content understanding is used in applications for improved content discovery and monetization. Val.ai powers visual search engine Valossa Deep Search and automated video metadata generator Valossa Deep Metadata.”

In addition to finding movies that slip your mind, Val.ai lets users browse broader topics to explore more options. Users can look up period films with battle scenes or romantic comedies with a strong female character. With the option to understand chained voice commands, the site can help people find their perfect content.

"Descriptive movie search is based on our research on what is called “Deep Content," shares the company on their website. "Deep Content is everything we can see and hear in a video, but cannot mechanically analyze – until now. Deep Content includes transcripts, audio, visual patterns and basically any form of data feed that describes the video content itself. After analyzing the deeper levels of the video, we automatically convert it into advanced metadata. This metadata is then processed by the beating heart of our engine: a cognitive machine learning system that understands natural language queries and matches it with our metadata."

The company hopes future versions will be compatible with voice command systems like Alexa, but until that becomes a reality, users can use the deep search feature online.

Join the Discussion
Top Stories