Frequently Asked Questions

Yes, the video feed of the existing installed cameras can be utilised depending upon the use case.

We support all cameras that are 2 MP or above with ONVIF. For RLVD/ANPR a camera with varifocal lens is required. The video feed for the camera can be provided through RTSP for H.264/H.265 stream.

Deep learning models, sometimes called “networks,” are the heart of a modern computer vision system. They take in image/video information, perform an inference and then output results. Each Pushak.AI application uses a specific model to return information about the contents from a video source.

Our models can be hosted on either one depending upon the requirements of the client

Yes, our models can operate on CPU. If you are looking for a real time analysis, then a GPU is recommended.

Our software is supported on Windows and Linux.

Traffic management systems, ITMS (Intelligent traffic management system) leverage AI to give data driven insights on traffic situations as well as automate fine generations for violations such as red light jumping, No helmet, triple riding, etc.

Pushpak ANPR provides accuracy 96%. The rate may depend on many factors such as image quality, weather condition, condition of the license plates, vehicle movement speed etc.

Yes, additional video analytics can be added. For example, RLVD setup can be combined with helmet detection, adaptive presence. Etc.

To continuously evaluate the efficiency of the city, smart cities use cameras and AI algorithms. Computer vision and other related technologies have significant advantages for governments. City managers can simply integrate and manage assets thanks to these technologies. Computer vision adds a layer of intelligence to cameras and help build cities of the future.