Video is the vastest information source. At the end of 2017 NVIDIA reported that there would be more that 1 billion of cameras around the globe by 2020.
The INSPEM start-up develops a video analytics platform, which would employ artificial intelligence and the blockchain technology.
The mechanics are very simple. On one hand, anyone on the planet Earth would «on a software level» be able to connect their surveillance camera located on the face of the building, balcony of their living quarters or at their own shop, private car, etc. All that is needed is the Internet access. The platform would be able to anchor the cameras to the GPS coordinates.
One the other hand, there would always be users who would require video footage containing the face of the person they need to find. The situations vary: it may be a missing relative, or an internationally wanted criminal. The user uploads a picture of a searched person and the algorithm starts to browse through the video cameras all around the world in search for them.
When the platform has recognized the face and found the video camera that caught that person into its sight, the system provides geolocation of that camera, the precise time, when the person was found, and the footage itself. Another party (the camera’s owner) receives a reward for the provided footage.
In that way, even the regular surveillance system of a living quarter or a commercial object may become the source of profit, since 99% of the recorded data would be of no use to the owner, but may be sold to somebody.
Given the colossal amount of the analyzed data, the system would employ miners’ capacities as the most effective and relatively cheap data processing mean. The employment of artificial intelligence would reduce the search time dramatically compared to the classic algorithms. Also, this technology has the «smart cities» establishment potential in collaboration with the world’s leading corporations.
The whole procedure takes place without revealing the personal data. The picture is uploaded to the system without the specification of the full name or other data. Likewise, the INSPEM provides only the GPS coordinates and the time, when the footage was taken. The system does not know any personal data. It only finds matches.
The protection from fraud schemes should be developed thoroughly. For each commenced search of a specific person a price must be payed, which would exclude the curious loiterers out. The system doesn’t provide access to the camera’s online translation. The video footage would only be provided if it contains the face of the person, whose picture was uploaded to the system.
INSPEM firstly aims on provision of city safety. For example, the wanted criminal may freely visit the grocery supermarket, get caught into the sight of the store’s surveillance cameras and freely go home, since this footage would remain within the store. By employing the INSPEM platform, the special services would be able to upload the criminal’s picture to the database, and any user video camera which recognizes this person would immediately receive a reward through smart-contract. It’s certain, that there are way more user and commercial surveillance systems, than any government control systems.
At this moment the start-up collects the Pre-ICO investments to realize the working model by the end of 2018 already.