Innovation That Matters

Motion capture | Photo source Shutterstock

Scientists teach computers to teach themselves

Telecommunications

AI scientists in China use self-learning to train a neural net to train itself

UNLOCK THIS INNOVATION AND MUCH MORE…

Become a member today and get early access to the ideas transforming our world from just £39 per month*

Exclusive member benefits:

  • Access to over 13,000 innovations
  • Monthly horizon scanning reports
  • Exclusive feature articles

Already a member? Sign in here

Spotted: Scientists at China’s Sun Yat-Sen University and Hong Kong Polytechnic University have developed a way for a neural network (a type of AI that mimics how a human brain works) to self-correct its understanding of a video of human movements by comparing its guesses to the guesses of other networks.

Up until now, to correctly identify human poses, neural networks required lots of data — cues about how a knee or elbow bends, and 2-D and 3-D renderings. The problem is that there isn’t always enough labelled data to feed a network. The researchers decided to demonstrate that a network could fine-tune its approximation of human movements by continually comparing the guesses of multiple networks with one another, and in this way teach itself.

The team included Keze Wang, Liang Lin, Chenhan Jiang, Chen Qian and Pengxu Wei. Qian works with the Chinese AI startup SenseTime.

Email: paadmin@polyu.edu.hk

Website: www.sysu.edu.cn, www.polyu.edu.hk

Contact: www.polyu.edu.hk/contact_us

Download PDF

Springwise Services:
Our expertise in spotting the latest innovations is the best resource to empower your team’s future planning.

Find out More