Colourful Past uses machine learning to instantly and intelligently recolour historical images. It applies deep learning to images from data sources across Australia and NZ. The AI is based on research published by a team from UC Berkley, trained to use educated guesses to apply colour to photographs.
We have combined traffic volume data with NASA satellite data to predict traffic-induced emissions across the Perth Metro region.
First, we use the traffic data from the Main Roads WA data on 988 road sections in the Perth Metro region. From this, we can see the traffic volume, speed, and type, such as freight or private vehicle. Then, we use modelling to predict the volume and speed of traffic on the rest of the roads in the Perth metro region: over 10,000 road segments in total.
Big data analysis has often been encouraged at GovHack events over the past few years, but finding the data and sanitizing in preparation for a hack typically consumes the majority of the weekend and leaves little time for creating a finalized product, hack or prototype. This makes us sad, especially as university students who often go through this mindless and unproductive leg work as part of the typical research process. We love the open data policy the Government is championing, but like wingy kids, we're always complaining for more. Yeh, sorry.
#BudgetHack is project undertaken by Perth Urbanist for GovHack 2016. The aim of our project is to use data from the Department of Local Government and Communities' MyCouncil and the Department of Industry, Innovation and Science's Region Innovation Data to explore corrolations between what local government spends its money on and the rate of return in terms of measures of innovation.