The possibilities of artificial intelligence and data analysis at your fingertips

Artifial intelligence based traffic congestion recognition. The road weather camera images of Finnish Transport Agency were fetched and the number of objects (cars etc.) was calculated using the neural network. The number of objects is visualized as a heat map on Google Maps. Try the demo

Easy-to-use image recognition demos that work in a web browser. The open sets are used to teach the neural network developed using Keras (library). The loading of the model and the prediction are established by Tensorflow.js.

Visualization demo application showing how to use deck.gl and react-map-gl to visualize LAM traffic data from liikennevirasto (Finnish Transport Agency) open data API.

The neural network is used to teach the self-driving car to drive independently on the video game tracks.

Neural networks can be used for creative purposes, such as synthesizing images from scratch. We explored generative models as a method of enhancing and drawing realistic human faces.

JAMK student team proposal to the Future City Challenge (FCC) in Jyväskylä. The aim is to embracing resource wisdom city.

Clickstreams, also known as clickpaths, are the route visitors choose when clicking or navigating through a web site. Using data analytics, we are examining the online shopping clickstream associated with the purchasing of products and sites from which Wikipedia has been accessed.

Purpose of this project was to do market basket analysis/research with common classification method (Apriori algorithm).

Object recognition using neural network either from the still images or video stream. Testing face recognition is integrated into the object recognition if the object is recognized as a person. Furthermore, the object recognition is integrated into the Nao robot.