IBM AI beats Microsoft's image recognition record

IBM's trained AI model achieved 33.8 per cent accuracy, beating the previous record of 29.8 per cent set by Microsoft in October 2014

IBM has claims to have beaten Microsoft's record for AI-based image recognition accuracy by developing distributed deep learning (DDL) software that can accurately identify images more than one-third of the time.

IBM said that, given any random image from a set of 7.5 million pictures from the ImageNet-22K database, its trained AI model achieved 33.8 per cent accuracy, beating the previous record of 29.8 per cent set by Microsoft back in October 2014.

What's more, IBM's system managed to achieve this in seven hours, while the process that allowed Microsoft to set the previous record took 10 days to complete.

According to IBM, this is a "milestone in making Deep Learning much more practical at scale—to train AI models using millions of photos, drawings or even medical images—by dramatically increasing the speed and making significant gains in image recognition accuracy possible as evidenced in IBM's initial results".

IBM smashed Microsoft's record by developing creating DDL software to help GPUs talk to each other. This was necessary due to the use of multiple servers with GPUs, with IBM fellow Hilary Hunt explaining that the more GPUs there are, or if they are of a higher quality, the learning time can actually be slower as they have more to 'talk' about.

"Basically, smarter and faster learners (the GPUs) need a better means of communicating, or they get out of sync and spend the majority of time waiting for each other's results," said Hunt. "So, you get no speedup-and potentially even degraded performance-from using more, faster-learning GPUs."

IBM's DDL software addresses that problem, and it should make it possible to run popular open source codes like Tensorflow and Caffee over massive neural networks and data sets with very high performance and accuracy.

"Our technology will enable other AI models trained for specific tasks, such as detecting cancer cells in medical images, to be much more accurate and trained and re-trained in hours."