Computer Vision and Maps for Self-driving cars
San Francisco, CA
Joined February 2018
- Tweets34
- Following58
- Followers207
- Likes1
Thanks for all your support over the last few years! We're proud to announce that we've been acquired! We will release more details over the next few months!
Ahead of the SF micro-mobility conference, a sneak peak at our vision-based sidewalk localization technology for scooters and e-bikes: youtu.be/NVCjbtLVnwk
Autonomy is an incredibly complex problem that can be broken into chunks. We have focused on just mapping for the last two years, and developed software that can robustly take just consumer-grade dashcam and GPS data and output beautiful lane-level maps at scale
"The most obvious bridge technology today, is the use of LIDAR for autonomous cars. Everyone agrees in principle that at some point in the future we’ll be able to make fully autonomous vehicles that drive with vision alone - after all, people drive without LIDAR." @benedictevans
Out of the hundreds of startups that pitched at #WebSummit2018, we are happy to have won first place for audience choice!! Check out @andrewkouri 's pitch and Q&A in front of the live audience! youtu.be/sn-_29bknz8?t=106
6
First exhibit of our new Road Quality algorithms, scaled across all of our Payver data
America's highways and roads are crumbling — here are the 10 states that have it the worst read.bi/2Q5TEOi
1
lvl5's 1st Data Science Meetup went well! Some cool new computer vision and data science concepts discussed in detail. Many more to come!
2
Ever wonder how Payver sees the road? lvl5's new techniques build 3D maps from just the camera footage of an iPhone. #hdmaps #payver #AutonomousVehicles
This heatmap shows depth inferred from context of a Payver video, just like your brain would do if you cover an eye.
We'll be looking for talent! nitter.gq/ycombinator/status/100…
1/ If you're an engineer, please come to YC headquarters on July 28 for our Work at a Startup expo. You'll meet the founders of 35 great YC companies who are hiring. Apply to attend: workatastartup.com/expo
Stay tuned for our first season of "The RADAR," a podcast featuring conversations with experts from the self-driving industry. Coming in July!
2
4
Demoware is gimmickware. Everything we're doing is designed from the beginning to scale. We could take shortcuts to make #maps faster, but it doesn't do anything to get #autonomous cars into mass-production. Maps have a very long tail and take a while to get right. #caveatemptor
1
3
Waze and other routing apps right now just figure out where there’s traffic. Not why. When you add cameras, the traffic problem becomes less heuristic based, and more about reasoning. AR combined with an HD map let’s you know how long until the traffic light ahead is green.
1
If there’s a collision blocking a lane on the highway, we automatically know why and merge car into another lane earlier to avoid last minute merge traffic jams
1
When at least 25% of vehicles on the road have a lvl5 dashcam installed, we will route vehicles based on real-time traffic light status. We can see that too many people are waiting at a red light and load balance across the city grid. #smartercity
The world is NOT flat! Around curves, our crowdsourced HD map helps you see through walls. Now, with improved 3D lane splines... beat that, LiDAR!
"In Keeney’s eyes, that’s another missed opportunity. 'That’s what they’re missing, and that’s what every other automaker is missing,' she says. 'Why has no one put sensors on their customer cars that collect data like Tesla has?'"
bit.ly/2F9dTof
1
It's actually really hard to properly compress, store, and send this amount of data. lvl5 is helping OEMs avoid common mistakes with #connectedcars
1
1