Computer Vision and Deep Learning Technologies
Computer Vision : Information extraction and classification from the contents of images and videos.
Image Processing: Analysis and processing of an image to produce its altered versions and to collect information about it.
Deep Learning : An extension of the Artifical Intelligence and Neural Networks technologies. It helps to detect objects from images by using multi layered deep networks imitating human brain model.
Smart solutions and projects based on image processing and recognition.
Utilizing the technological transformation in Computer Vision and Deep Learning and applying this transformation to the applicable domains wherever the smart systems can be used.
Having based on Smartvist Framework and SmartVist FrontEnd, SmartCity product family and SmartCloud services can be used for smart applications as they are making our lifes easier and they are becoming part of our daily lives.
Smartvist solutions can be used for smart cities, security systems, video analysis, industrial automations, autonomous vehicles and medical image analysis.
Face Recognition : It is the most successful usage area of Deep Learning technologies. Most successful Face Recognition usage examples are implemented via Convolution Neural Networks within the last three years.
It is possible to recognize all kinds of faces when the right number of sample images containing faces. Face Recognition solutions are expected to be heavily used in the smart city solutions, surveilance and security, marketing and entertainment sectors.
Vehicle and Person Detection : Detecting vehicle and person are very important applications in drones, self-driving cars and security domains. Advancements in the Convolution Neural Networks (CNN) improved the accuracy and success rates of detections.
Nowadays some car makers are using the customized CNN solutions. It is expected that the use of CNN applications with mobile phones will further increase and CNN technologies will become more prominent.
Object Recognition : Thanks to Convolutional Neural Network technology and Image-net contests since 2010, the success rate of object recognition processes reached to %95 levels. The success rate is measured using 1000 objects from the 150.000 randomly chosen image and video.
"letgo" like e-commerce web sites are able to operate with the minimum number of people because of the Object Recognizing system. This technology hasn't been widely used in the industry yet but it is expected that there will be successful applications in security, defense, transportation and e-commerce sectors within very short time of period.