After working (furiously) for the past few months, our team is very excited to make public a limited set of demos that show our VideoRecon analytics platform in action!
For those of you unfamiliar, VideoRecon is a platform that we built that uses deep learning in order to “watch” and “listen” to videos and return metadata “auto-tags” based on objects, themes, and descriptions it identifies in the video content, so that the video file itself can then be more easily searched for, filtered for, and used in a wider context for analytics.
Without these metadata “auto-tags” in place, it would be very difficult to use video effectively in this context, unless “manual” tags were applied to each video by a human. As many in the ecosystem have seen, the process of manual tagging has a lot of potential issues, including the length of time needed by a human to actually watch and tag each video, the lack of objectivity between human taggers (i.e. bias), and the sheer numbers of people needed to manually tag videos to meet designated timelines and business objectives.
VideoRecon solves all these challenges by simply allowing a user to run their videos through our platform, lean on our deep learning algorithms to objectively evaluate each video and return auto-tags in a fraction of the time required by a human equivalent to actually watch a video. This enables users to more accurately tag videos and have them available for search/filtering/analytics much more quickly.
Having the tags in place will then enable enterprises and users to find and get the most targeted video content in front of clients, employees, or other business stakeholders faster and more effectively than ever before. Examples of this might include marketing or support videos posted online for customers, internal videos used for training or communications, or even a “digital experience” component of a traditional product that incorporates video.
Our demos can be accessed at videorecon.io and is the first glimpse into the full VideoRecon platform that we’ll be rolling out in phases over the coming months.
Among our plans, we’re planning on introducing an upload portal to allow videos to be manually uploaded by a human operator, a developer-friendly API to enable VideoRecon to easily be tucked into the applications and workflows of our users, and integrations with other popular third-party services.
In short, our goal is to make it as easy as possible for our users to take advantage of the power of deep learning for video tagging/analytics, no matter if you’re a professional video tagger, business analyst, developer, or someone who just needs more insights out of the mountains of videos stored on your company’s servers/storage or cloud storage repositories.
Please let us know if you have any feedback or would like to run your use case by us! We’re constantly improving our platform and would love to know what you think!