Artificial Intelligence (or AI in short) is a well-known term in the world of modern computer science. AI systems are used to simplify human lives where computers can perform human-like tasks without the intervention of any human operator. A typical example of AI is computer vision. Computer vision applications can process image or videos and are able to do various tasks along with detecting and recognizing faces, detecting moving objects, count people in a busy street or shopping mall and so on.

What we do

The primary target of AI applications is making a computer learn to perform tasks which needs human resources and their intelligence. Algonics is exploring different fields where AI can be applied to solve real life problems. So far, Algonics is applying AI in the fields like Data Analysis, Remote Sensor Monitoring, Image Processing and Computer Vision.

We have recently developed an intelligent driver monitoring system, where an Embedded Linux Board connected with a camera can detect whether a driver is tired or distracted while driving a car. We are also applying AI and deep learning technologies for face recognition, gender & age detection, behaviour prediction and crowd counting.

How we do

We mostly use open source technologies for our development. The backbone of most of our AI applications is Python, sometimes combined with C++. Open source tools like Google’s Tensorflow, dlib library, OpenCV and Keras are widely used along with custom-made training models and algorithms.

With the help distributed task management systems, distributed file systems and highly available cache servers, we are able to develop high-load image processing systems which can process live streams from multiple cameras 24x7. We can also develop Image Processing and Deep Learning systems on low-end hardware like Linux based embedded boards and SBCs, which can be used in small areas for computer vision applications.

To view our projects on AI, Click Here.

Read this article to know how we integrate Android, Linux and Cloud backend for Face Recognition systems.