A long video of us operating our truck without a driver, hauling cargo from Dallas to Houston

Been working on self driving for years now and we finally started rolling our trucks with no driver!

It was fun talking to Ben and Michael on the Adventures in Machine Learning podcast.

To showcase our industry-leading approach to developing and deploying autonomous technology, we've created a video series demonstrating how the Aurora Driver makes safe decisions about real-world situations. Check it out here:

dlib 19.24 is officially out. Some quality of life improvements and deep learning stuff. Changes at

We've raised a $8.3m war chest to reinvent the paradigm for data privacy on the internet. While we build towards moonshot goals, we ship simple, powerful APIs so developers can build incredible user experiences today. I'm deeply grateful for our team and community. Thank you!

Next dlib release is out, some cool new deep learning stuff in here thanks to @archdria. Check out

Merry Christmas! We got you a new telescope. The James Webb Space Telescope launched today, beginning a one-million-mile journey to see 13.5 billion years into the past. Follow @NASAWebb and join the quest to #UnfoldTheUniverse: https://t.co/TlYpoUHdJu

The Aurora Driver: Autonomously Driving Over 200 Miles

Pretty sweet spotlight from @CBSNews on @aurora_inno! https://t.co/yfT1wYFEhK

Our truck doing its thing 😁

If you review code on GitHub and wish you could comment on *all the lines* go upvote this request to make the GitHub UI allow it :D

dlib 19.22 is officially released. Check out https://t.co/YpXr6h04xg to see the list of improvements :)

Just tagged the dlib 19.21 release. Main change is support for CUDA 11.0, cuDNN 8, and better build experience for python users on Windows. Notes at

A cool summary of some of the tools in dlib with links to articles that elaborate on what you can do with those features.

dlib 19.20 is out. Supports CUDA 10.2 and the latest OpenCV. There are a number of other improvements and new docs like a DCGAN example, courtesy of Adrià Arrufat (see https://t.co/YpXr6h04xg for more details)

New tutorial!🚀 Automatically find optimal hyperparameters using dlib's "find_min_global" function: - Parameter free and provably correct - Easy as a single function call - Guaranteed global min https://t.co/xKsQEd8ERp 👍 #Python #ComputerVision #MachineLearning #DeepLearning

New tutorial!🚀 How to tune dlib shape predictor hyperparameters to balance: 1. Speed 2. Accuracy 3. Model size Full tutorial: https://t.co/WIZPAFXgFD 👍 #ComputerVision #MachineLearning #DeepLearning #ArtificialIntelligence #DataScience #Python CC: @nulhom (creator of dlib)

New tutorial!🚀 Learn how to train your own custom Dlib shape predictor. - Step-by-step tutorial - Full code walkthrough - Includes trained models https://t.co/ATLlEAa4lQ 👍 #Python #MachineLearning #ComputerVision #DeepLearning #AI #DataScience CC: @nulhom (creator of dlib)

dlib 19.19 is out. See https://t.co/YpXr6h04xg for details. One cool thing is the global optimizer find_min_global() is now nice and speedy for a wider class of functions. It still works as described here https://t.co/Ws5cf1nln8 but now autotunes itself to the objective function.