Today, we are announcing the v1 launch of the Nuron.World API. NuronLabs was founded with the goal of enabling truly safe & reliable real-world neural networks in production applications and this milestone brings us one step closer to achieving that goal. A well designed API is essential because it allows for deeper integration into the production application stack. This unlocks the ability to learn from and adapt to data automatically resulting in greatly compounded value over time. Most importantly, this allows real-time safety monitoring to verify that the neural networks are performing reliably.

What you can do with the Nuron.World API

There are many components to ensuring safe…

Deep learning research done wrong can be extremely expensive in terms of computation and time. In addition, it could also hurt the environment. A poorly structured deep learning project might involve:

  1. Write some code.
  2. Start training.
  3. Wait for something to fail.
  4. Fix and repeat.

This process slows down progress and wastes resources. However, this process is often hard to break as a developer because it feels productive. The dopamine hit from solving an issue is addictive and the time spent on waiting for the model to finish training feels like work. …

Have you wondered why self-driving car projections keep getting pushed back from 2018 to 2019 to ‘way in the future’ to ‘may never happen’? While self-driving cars prototypes today make for good demos, they are not safe enough to deploy in the real world without substantial human intervention.

One of the main impediments of self-driving car safety is the phenomenon of distribution-shift. This is related to the long-tail problem that’s oft discussed but is not quite the same. Distribution-shift occurs when the real-world data seen by a model during inference is different from that which it was trained/evaluated on. …

Although the Garter hype cycle believes that we’ve reached peak Deep Learning in 2019, I believe Deep Learning is in its early stages and there are numerous extremely valuable opportunities that lie ahead. Here are some of the opportunities I see for the next decade.

1. Real-world deep learning

The key shift that will happen this decade is the movement of Deep Learning from the web to the real world. While the highlight of Deep Learning in the 2010s was classifying YouTube videos of cats, this decade will be dominated by autonomous vehicles, augmented reality, and other real-world applications.

While the previous generation of…

Deep learning has seen a rapid increase in popularity over the last 5 years.

Given that its a field in the rapid stages of its development, it has brought about an influx of researchers and institutes. This surge was so rapid that in 2018, tickets for a niche machine learning conference, NIPS (now NeurIPS), sold out in less than 12 minutes.

This article is the start of a series that will provide some background for the problems NuronLabs provides solutions to.

Precision-Recall is an important metric that measures performance on a few tough visual perception problems such as Object Detection, and Instance/Semantic Segmentation. Models trained to solve these tasks are often sensitive and perform very differently on different visual distributions.

Thus, when we compare models, it is useful to have a more fine-grained, yet holistic view of their performance characteristics. To this end, precision-recall turns out to be a fairly useful metric to track when considering where the model might fail.


Building a neural layer for reliable real-world deep learning

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store