Artificial intelligence is a tool that has become increasingly necessary nowadays, but it can only be created and enhanced through the accurate analysis of data.
In 2015, The University of Oulu in Finland started creating a project named Valossa, an AI company that came up with the idea to use the technology to automatically understand emotions, actions, people, speech, sound and visual concepts in video format for the benefit of the media and entertainment industry.
The company has been expanded to New York, USA, and they define themselves a “a set of surgical knives for video intelligence”.
The main goal focus is deliver the most comprehensive video-understanding AI. and with the Video Insight Tools Preview program, any vide professional can search functionalities and a way to tag and train the AI. The content-understanding AI is also enhanced with novel tools that facilitate scene-level video search and insightful reports that unveil what’s truly relevant inside the video library.
Thus, the tool can help you identify all sorts of videos. For example, one of women who are laughing, or all the scenes where Liam Neeson jumps over a car or is looking for somebody.
The way that it works is, basically, the AI goes through your video files and indexes every recognized person, object, scene and specific sound. Simply describe what you’re looking for and the search engine will quickly find the best matching clips from your videos. This will save you time when you’re looking for specific concepts within your content.
Valossa is able to examine the details and find what’s important in any video. Overviews summarize which recognition categories are dominating in the video and what are the most relevant names appearing in the audio and visual track. The heatmap view visualizes everything that the AI has detected from the content.
This tool is, at the same time, trainable. That means you can add more information to the AI software and it will be able to create customized face galleries quickly and easily. Simply name the detected unknown people and add them to the AI training process. The feature begins with tagging enabled and evolves towards training features during the program
This plataform has a plan to keep growing and it expects to mature the tool using weekly feedback from the users.