Canadian startup applies AI to deep neural networks to improve efficiency and transparency
Canadian startup DarwinAI has emerged from stealth mode to offer its next generation technologies to artificial intelligence developers to help reduce the complexity and guesswork involved in designing efficient, high performance deep neural networks.
The company provides its Generative Synthesis platform to apply AI to understand how neural networks work. It generates a compact network to maintain functionality and provide explanations for its predictions, which the company says is particularly important for companies in regulated industries. DarwinAI says its technology has generated a deep neural network 4.5 times more computationally efficient than the one produced by Google’s AutoML and Learn2Compress platform for image classification, as well as an optimized version of Nvidia’s DetectNet object detection network that is 12 times smaller and 4 times faster than the original.
DarwinAI has had success with proofs-of-concept for automotive, aerospace, and consumer electronics clients, and its technology is particularly applicable for edge implementations, according to the announcement.
It has raised $3 million in seed funding, which was co-led by Obvious Ventures, iNovia Capital, and angel investors from the Creative Destruction Lab accelerator in Toronto.
“From autonomous vehicles to mobile devices, we are seeing edge-based scenarios where AI is having a profound impact on business outcomes,” said Sheldon Fernandez, CEO of DarwinAI. “A critical challenge in this realm is designing these powerful networks to run in situations where computational and energy resources are limited. To this end, our platform is much more than an optimization tool. It allows engineers to collaborate with powerful AI to develop efficient and interpretable network models.”
DarwinAI was founded by renowned University of Waterloo alumni including Dr. Alexander Wong, a professor at the university, Fernandez, who co-founded software consultancy Infusion, which was acquired by Avanade last year, and Arif Virani, a former consultant for McKinsey & Company.
Wong’s team presented two award-winning papers in workshops at the Neural Information Processing System (NIPS) conference in 2016 and 2017, and he is a founding member of Waterloo’s AI Institute and Canada’s Research Chair in AI.
“I’ve been spending time with development teams at some of the world’s leading tech companies to learn about the state of AI – and all of their input leads back to Darwin,” said Nan Li, Managing Director of Obvious Ventures, who co-led the seed investment in DarwinAI. “The company’s product accelerates deep learning development in two fundamentally critical areas: by enabling teams to quickly optimize neural networks design, and by shedding light on the inner workings of such networks to inform faster training and better model development.”
Several tech giants have released tools to improve the efficiency and transparency of AI systems within the past year, including IBM and NEC. A key difference is that DarwinAI’s Generative Synthesis platform actually uses AI to create new networks to address user requirements.