Cortana may eventually be able to peek inside your refrigerator following news that Microsoft is bringing its machine learning tech to kitchens. The computer firm is working on a new computer vision module called SmartDeviceBox, which will be installed inside fridges to make them capable of recognising objects inside and telling owners when they need to re-stock.

The SmartDeviceBox is essentially an internet-connected camera unit that sits inside fridges and tracks the objects inside. The unit contains an image processing system based on Microsoft's suite of deep learning tools, which has been trained to recognise a variety of products such as milk cartons, ketchup bottles and pickle jars. The system is able to collate these into an inventory list that customers can view in list form using a smartphone app.

SmartDeviceBox image recognition
The unit uses image recognition and deep learning to track products inside the fridge Microsoft

Although the SmartDeviceBox is still in the prototype stage, Microsoft believes that the technology will eventually be capable of helping customers keep better track of what's in their fridge and notifying them when things need replenishing, ultimately meaning less waste and more efficient shopping trips.

The first of such units will be installed in fridges from Swiss manufacturer Liebherr, who is collaborating with Microsoft on the initiative.

TJ Hazen, Principal Data Scientist Manager at Microsoft, said: "Although the current system is a prototype, the deep learning technology it uses is already very powerful and rapidly maturing. Liebherr and Microsoft are actively improving these technologies to turn the newest generation of refrigerators into smart appliances that don't just cool your food but interactively help you with your food management."