Self-driving for everyone
Visioner is a Deep Learning mobile app, that can connect to your car via our hardware and controll it using only the video stream from your phone camera.
Watch the video >Works with a push of a button. Visioner is going to engage automatically after the turn on and will perform acceleration/deceleration and steering, analyzing the video and passing the model's output to a vehicle.
Our state of the art model can detect objects and evaluate distance to the object using triangular similarity algorithm and Deep Learning
Computer vision detects lanes and evaluates offset of the vehicle. The model curently needs to be at a lane road to fully function.
Every car with ACC (Adaptive Cruise Control) system has a set of ECU's to electronically controll it, as known as CAN. Our hardware can simulate those signals to read and write to your car to get outputs from the model straight to the car.
I'm a software engineer with more than 5 years of experience in intelligent software systems applications. Visioner has data of more than 10,000 miles recording. Using that data we trained the computer vision and the deep learning models, combining both knowlage-driven and data-driven approches for best accuracy.