Face Detection and Recognition System

Face detection and recognition system for security. Able to recognize the person and make a decision whether he is allowed to be in a certain area and is it needed to open a door. Face detection systems made by ArtiD are highly reliable and fast performing tools for a variety of implementations.

aliusenergyel841-fullscreen.jpg

Solar Panel Defect Detection

Solar panels during production can have a variety of defects like scratches, cracks and many others. Our algorithm provided the possibility to autonomously detect such defects during the production. Deep Learning solutions are able to detect 95% of the defects during production line.

Face Scanning System

iOS system developed to build a highly accurate 3D model with the possibility to save in .obj with texture or .ply using depth data from iPhone front camera with infrared sensor with 30K referring points. A 3D scanner can be used in many areas from fashion to the game industry. 

Hand Segmentation and Gesture Recognition System

This Deep Learning solution was developed for real-time segmentation of a hand in a video stream and also gesture recognition for remote control of devices.

GAN-Based Molecule Docking Algorithm

Molecular docking problem is a very important problem nowadays as it is required for drugs discovery. The project was aimed at creating a new way of docking simulation in a new and much faster way. Using generative adversarial network let us create a new type of algorithm which was 10X faster than the original ones.

Genome Analysis

Genome analysis is a hard problem as it requires huge data analysis and machine learning capabilities. Analysis and prediction by time series and convolutional neural networks on genomic data.

Anomaly Detection System

Anomaly detection system for a chemical factory was developed by our team and was able to prevent and detect possible failures during the factory process. Historic data from 21 sensors were processed and used to train the AI model System became able to predict future failures 39-44 hours in advance 99.782 % accuracy achieved as tested. Future development of a system allowed us to develop a fully autonomous anomaly detection system where user can train on any new data a new model for the same anomaly detection and prediction.

NLP Based Contract Analysis System

Using the newest NLP models we were able to solve the problem of a huge number of companies. Lots of companies hire huge offices of people to analyse the contracts they have and receive. With our system, you can get insights from every contract you have in minutes without any human interaction.