Congrats on the launch! The constraint of a ~60W compute budget for fully onboard autonomy in a GPS-denied environment is impressive, I'm curious about the 'feature-poor' aspect you mentioned. Warehouses are notorious for having repetitive scenery which usually hard for visual SLAM regarding loop closures. How do you prevent the drone from getting lost when aisle 4 looks exactly like aisle 40? Best of luck!

I checked the page and it seemed really cool, would like to give it a try!

Hey sorry for the late reply - just email me at hello@rezi.io

In for English but for a friend of mine.

Start with this course: https://www.cs.cmu.edu/~16385/ Then you can check udacity and coursera courses for DL. Try not to go directly into DL but first understand basic computer vision terms.

It seems very interesting! What about support for pytorch models or .onnx? I usually use the pytorch->onnx->tensorrt to deploy models.

Hi! We do not currently have PyTorch model but it's one of the next items in our roadmap.

If you have a more custom pipeline, we have a custom environment where you can deploy any custom code with specific package versions!

Awesome website! Great work, it would be nice to have a sort parameter to sort by score or price.

Thanks! Yes, I really need to add sorting... I'll do that next, thank you.

Very nice build indeed! I was comparing it with Lambdas build and it's pretty good on comparison, any thoughts on the Threadripper vs Intel? Also regarding the PSU, I've seen some pretty nasty reviews on that one (https://www.amazon.com/Rosewill-HERCULES-1600-Continuous-Has...). How did yours perform so far?

Pretty nice, I'm really interested in pipelines for deep learning at scale, they link to this article (https://blogs.dropbox.com/tech/2017/04/creating-a-modern-ocr...), if anyone has some other insights I'd be thankful. So far I used TensorRT to deploy to inference servers. But there is a lot of boilerplate and something like a 'load balancer' for DL networks would be very interesting.

Awesome work, I'd try to check it out once it's up again.

The whole ROS ecosystem is huge, most of research platform robots are ROS enabled. Check the PR2 (http://wiki.ros.org/Robots/PR2), Turtlebot (https://www.turtlebot.com/), the Nao series (http://wiki.ros.org/nao), even Boston Dynamic's Atlas robot (http://gazebosim.org/tutorials?tut=drcsim_atlas_siminterface...) or Nasa's Robonaut (http://wiki.ros.org/Robots/Robonaut2).

It facilitates work by not having to rewrite a bunch of code every time. Also, most robotic related libraries like OMP, or TF are supported.