Neural networks are used in multiple projects completed by ZUZEX developers.
Our developers have completed multiple tasks related to implementation of neural networks.
Depending on a task different structures of neural networks were used:
- one-layer and multi-layer perceptrons for data classification and pattern recognition,
- multi-layer neural networks and rbf networks (single and multil-layer) for prediction and function approximation.
We have knowledge and experience of selecting an optimal structure of a neural network and parameters of neural network layers by using cross validation based on various metrics.
Mechanism of neural networks was implemented on multiple programming languages, including Matlab, R and C++, both from scratch and using special libraries and packages.
Virtual try-on tool for makeup products
The system consists of a face detection and segmentation model and an algorithm that allows recoloring objects without losing their original texture.
Online sign language interpreter
AI algorithm that converts video of a person using sign language into a text transcript
Workout helper app
Mobile app for the estimation of proper body positions during the workout.