Earlier this year, researchers at Google Brain announced the creation of AutoML, an artificial intelligence that’s capable of compiling its own AIs. They decided to present AutoML with its biggest challenge to date, and the AI that can create a ‘child’ program, made one that outperformed all of its human-made equivalents.
The result? AI-ception!? Google’s researchers automated the design of machine learning models using an approach called reinforcement learning. AutoML acts as a controller neural network that develops a child AI network for a specific task.
For this particular child AI, which the researchers called NASNet, the task was recognizing objects – people, cars, traffic lights, handbags, backpacks, etc. – in a video in real-time.
The developer AI (parent) would evaluate NASNet’s performance and use that information to improve its child AI, repeating the process thousands of times.
When tested on the ImageNet image classification and COCO object detection data sets, which the Google researchers call “two of the most respected large-scale academic data sets in computer vision,” NASNet outperformed all other computer vision systems. Most interestingly, all of the others were man-made, not AI generated.
According to the researchers, NASNet was 82.7 percent accurate at predicting images on ImageNet’s validation set. This is 1.2 percent better than any previously published results, and the system is also 4 percent more efficient, with a 43.1 percent mean Average Precision (mAP).
Additionally, a less computationally demanding version of NASNet outperformed the best similarly sized models for mobile platforms by 3.1 percent.
We have seen other computer-generated structures that theoretically outperform humans – most famous examples are the Wendelstein 7-X stellarator, airplane parts such as wing supports, turbine propellers etc. But those didn’t involve machine learning, or learning in general and could not be called AI-generated.
As for NASNet specifically, accurate, efficient computer vision algorithms are highly sought after due to the number of potential applications. They could be used to create sophisticated, smart robots or to help visually impaired people regain sight, as one researcher pointed out.
The Google researchers acknowledge that NASNet could prove useful for a wide range of applications and have open-sourced the AI for inference on image classification and detection of objects.
“We hope that the larger machine learning community will be able to build on these models to address multitudes of computer vision problems we have not yet imagined,” they wrote in their blog post.