Start your future project with a versatile and skilled engineering, R&D team
One of the Major challenges in surveillence is detecting and tracking people. We intend to solve the problem by using deep learning in real time and improving the current state of art.
(GANs) are a way to make generative models by having two neural networks(Discriminator and Geneartor) compete with each other. At MAIVRIK we are training Generators to output images, faces etc.
We are exploiting use of Deep Learning and Soft computing principles to develop various applications ranging from people detection, vehicle detection, classification, counting, tracking, survillence.
Shadow detection is one of the major challenges in vehicular classification and tracking. We are working on the use of deep learning to remove shadows from images and moving vehicles.
We have used Aritifial Neural Nets (ANN) in our attempt to solve the problem of network intrusion detection. Detecting a malicious entry or an intruder is the cornerstone for any cyber security threat detection system.
We are researching on using efficient machine learning techniques to solve multifarious problems in cyber security. Another largely unsolved problem is identification of a compromised social media account.
Sentiment Analysis is seen as domain dependent problem and performance of such systems is therefore highly dependent on the amount and types of data available in individual domains. In order to overcome such problems we tend to develop a novel system utilizing variety of machine learning algorithms for multiple domains.