Enhanced search and reporting features for web-based AI video insight platform

Screen shot 2018 01 08 at 14.18.54

Artificial Intelligence (AI) company Valossa has updated its video insight tool Valossa AI, which detects and identifies people and ‘visual concepts’ as well as speech and sounds, with features including scene-level searching, video overviews, new reports and custom training.

With the AI video analysis space getting increasingly crowded with competing products and services, Valossa says its updates are part of a drive to ensure it is delivering the most comprehensive video understanding AI tools available.

Valossa AI’s updated search function enhances the way you can search within your video clips. The AI goes through your video files and indexes every recognised person (see below), object, scene and audio sound, so you can search for “Audrey and Cary in a car”, for example, and it will show you all your matching clips.

Screen shot 2018 01 08 at 14.16.09

The reporting of the data it gathers about your clips has also been refreshed, with new overviews summarising which recognition categories are most prominent in your video, and which are the most relevant names appearing in the audio and visual track.

Screen shot 2018 01 08 at 14.19.46

A heat map view shows everything the AI has detected from your content.

Another update to Valossa AI is a ‘tag and train’ tool, making it possible to customise the AI to match your requirements. So you can create personalised face galleries by naming any unknown people Valossa AI detects and adding them to the training process.

Screen shot 2018 01 08 at 14.22.14

Other changes to the software includes an improved video upload feature and a new pay-as-you-go usage model.

Valossa provides an initial free quota that’s currently in excess of 300 minutes for trying out the system.

It is available as an online platform or via APIs, to integrate into existing systems. You can find out more at https://valossa.com/

Screen shot 2018 01 08 at 14.18.04