In pictures: Microsoft's Imagine Cup finals

The projects fall into the categories of software, cloud computing and game design, with an emphasis on solving some of the world’s most pressing problems.

This year's worldwide finals were held in Sydney, where more than 350 contestants from 75 countries converged for the final leg of the race.

Team quadSquad: Computer Academy "Step" Donetsk, Ukraine

"The gloves capture the hand movement pattern and finger gestures," explained team member Maxim Osika. "The data is transmitted to the mobile device where the analysis happens ... the application on the smartphone is then able to produce the sound for those signs."

One of the benefits of the system is its flexibility: because the software is user-taught, the type of gestures or signs are easily defined. This means the user does not need to be proficient in traditional sign language.

Mobile Eye: Auckland University of Technology, New Zealand

During a demo, the app correctly read our name tag aloud and was able to identify a large spectrum of different colours.

Mobile Eye (cont.)

"It uses a mixture of artificial intelligence techniques like finding colours, reading text and so on. But it also uses human intelligence to find context-specific information – the app can route an image to your family or friend on Facebook to answer your queries.

This enables the blind person to go shopping independently or find out about objects on their own just using their mobile phone."

Mobile Eye (cont.)

wi-GO: Universidade da Beira Interior, Portugal

It is primarily designed for shoppers who cannot maneuver a cart unaided, such as people in wheelchairs.

wi-GO (cont.)

"The Kinect can detect a human skeleton," team member Ana Figueira said. "So it will lock onto a person and if they wave once it will just start following. They don't need to wear any kind of tag: it just works."

The contraption has additional rear-facing motion detectors as a safety precaution.

Dancing Pillow: Yarmouk University, Jordan University of Science and Technology, Jordan

It is designed for hearing-impaired individuals (and presumably very heavy sleepers).

Team Symbiosis: Aristotle University of Thessaloniki, Greece

"Symbiosis is based on trying to normalise the environment to the patient's needs," said team member Leontios Hadjileontiadis.

"We try to facilitate everyday life. For example, instead of writing with sticky notes, they can use sunglasses with augmented reality in front of the screen."

The product utilises the patient's mobile phone to help them remember faces (via photos) and jog their memory. The app also sets up a direct link to the caretaker's phone with emergency alerts, which allows the patient to be more independent.

The project was developed using Kinect for Xbox 360, Windows 7 and Windows 8.

Team uCHAMPsys: National Cheng Kung University, Taiwan

Team uCHAMPsys (cont.)

The system will then offer suggestions on exercises and activities to help them lead better lifestyles. The data it analyses covers a wide range of topics, ranging from energy consumption to sleep time management.

Team Coccolo: Tokyo National College of Technology, Japan

This green-friendly solution, which was developed using Windows Azure, Silverlight and Visual Studio, enables LED lamps to communicate with each other and automatically dim when there is more light in the room than needed.

Team Coccolo were grand finalists in the Microsoft Imagine Cup 2012 in the category of Software Design.

Team StethoCloud: University of Melbourne, Australia

"We basically took a stethoscope head, ripped out the body and added a microphone inside," explained team member Hon Wen Chong. "It's actually quite cheap. The whole thing only costs around $15 to $20 to make."

Team StethoCloud then built an application that prompts the user to enter personal data such as their date of birth and questions relating to symptoms.

"It asks the important questions that a doctor would also ask," explains Hon Wen.

The user then uses the 'Stethomic' to take five-second recordings of specific areas of their body. The audio files are then uploaded to the cloud where they undergo a diagnostic algorithm. The results are then sent back to the user, together with a treatment plan.

Team StethoCloud (cont.)

Let IT Bee: Kookmin University, Sangmyung University, Sungkonghoe University, Korea