EU startup ENOT has launched what it says is a user-friendly universal framework for a wide range of neural network architectures, delivering ultra-fast results, already used by 20+ enterprises, including Weedbot, LG, Huawei, Dscribe, Hive.aero, PicsArt and more.
Latvia-based ENOT styles itself “a leading developer of neural network optimisation tools,” and has released its AI-driven optimisation technology for deep neural networks, used by AI developers and edge AI companies, and raised undisclosed early seed investment from New Nordic Venture.
The company says the integration of its framework makes deep neural networks faster, smaller and more energy-efficient – from cloud to edge computing, with the tech promising to “help achieve outstanding optimisation ratios resulting in acceleration up to 20x and compression up to 25x, thus reducing total computing resource (hardware) costs by as much as 70%.”
ENOT explains it applies a unique neural architecture selection technology that outperforms all known methods for neural network compression and acceleration, and that its framework has a simple to use Python API that can be quickly and easily integrated within various neural network training pipelines to greatly accelerate and compress neural networks.
|
|
Its framework takes a trained neural network as input, after which it selects a sub-network that has the lowest latency and can ensure no accuracy degradation. The framework has a python API that can be quickly and easily integrated within various neural network training pipelines.
W’re also told it also allows users to automate the search for the optimal neural network architecture, taking into account latency, RAM and model size constraints for different hardware and software platforms.
In addition, ENOT says its neural network architecture search engine technology allows users to automatically find the best possible architecture from millions of available options, taking into account several parameters such as:
- input resolution,
- depth of neural network,
- operation type,
- activation type,
- number of neurons at each layer, and
- bit width for target hardware platform for NN inference.
Through all of this, ENOT says this technology helps customers achieve significant efficiency through cost savings and much faster product launches, and can reduce time to market.
The ENOT framework is aimed at companies that utilise neural networks on edge devices, such as companies in the following fields:
- Electronics
- Healthcare
- Oil and gas
- Autonomous driving
- Cloud computing
- Telecom
- Mobile apps
- Internet of Things (IoT)
- Robotics
That’s a very wide range of potential customers in that list, so the potential is clearly huge, with ENOT also announcing it has already successfully completed more than 20 pilot projects where it optimized neural networks for several leading tech giants from around the world, among a dozen medium-sized OEMs and AI companies.
Just some of the results in the cases and industries ENOT serves are as follows:
- ENOT delivered 13.3x acceleration of a neural network used by a top-three smartphone manufacturer as part of the image enhancement process. The optimization reduced the neural network depth from 16 to 11 and reduced the input resolution from 224×224 pixels to 96×96 pixels, yet there was practically zero loss of accuracy.
- Another project with the same company delivered 5.1x acceleration for a photo de-noiseing neural network – again with an almost imperceptible change in image quality – even though the network had already been manually optimised. For end-users, this translates to faster processing and significantly extending battery life.
Pricing is US $45, $55 or $60 per month, with details here, and there is a two week free trial.
Sergey Aliamkin, CEO and founder of ENOT commented, “Today, when neural networks are widely used in production and applications. Neural networks should be more effective in terms of consumption of computational resources and affordable. Their implementation should be faster, better and cheaper for all.
“ENOT is at the forefront of next-level AI optimization, helping bring fast, real-time levels of AI advancement through high throughput data into reality as science fact, and our journey has only just begun with examples such as the Weedbot laser weeding machine that gained a 2.7 times thanks to the ENOT framework.
“At ENOT, we’re not resting on our laurels, but shooting for the stars, and as always, the best is yet to come.”
ENOT reminds us it was founded in 2021, and that its team comprises seven engineers, among whom are several PhDs in computer science, and is three-time winners of the global Low Power Image Recognition Challenge, outperforming the MIT, Qualcomm, Amazon and other teams, which is pretty impressive!
AI can’t just be about dystopian nightmares where Skynet, Terminators, and other rogue AI’s just take control. Those science fiction movies aren’t just thought-provoking entertainment, they are also warnings not to let AI take humanity over and for those dystopian realities come true, but to make sure AI helps us on our terms, without dominating or destroying us, and without some loophole that turns the AI evil.
All of today’s technologies, including the achievements of ENOT, are all the technological ancestors of future AIs to come, so as a human race, let’s do what we can to make sure technology stays a growing friend, rather than a fatal frenemy, and Godspeed in ENOT’s quest to truly democratise AI for all to the betterment of human life in the best of all possible ways.
Until then, if your company sounds like it could use apply ENOT technologies, a part of the already invented future just got a bit more widely distributed, and now you know about it, too.
Whether you can use it or not, or tell someone else about it, is up to you, but unless humanity wipes itself out in a nuclear, financial, medical or environmental catastrophe first, we’re on an upward trajectory where, as I always like to say, despite the challenges we face, and thus add to Serget Aliamkin’s sentiments, the best truly is yet to come, and all of our actions are the steps forward we’re all taking to get there.
I just hope future AI’s truly are our friends, and not our overlords, so ENOT, AI and all tech companies of the world – humanity is counting on you to get it right, and to do the right thing.
ENOT alone is not enough for that far off future, but it sure sounds like a good news story for once in 2022 after COVID, inflation, the ongoing horror of new war, inflation and more, with humans still working to advance the state-of-the-art and deliver tangible benefits to others, and eventually upgrade everyone’s experience.
Long may it continue, and it will be fascinating to watch ENOT’s progress – and those of its competitors and contemporaries, because despite the challenges we all face, we do live in very intresting times.
