After my last video a lot of people suggested to train an AI to parallel park.
So here it finally is: AI Learns to parallel park!
https://t.co/N1BgH7WKqr
Really enjoyed working with the latest @unity3d ML-Agents!
#madewithunity#MachineLearning#ReinforcementLearning#AI
@Rainmaker1973 Was always wondering if my video would ever end up in your posts๐
Love your content! Been a follower for a long while.
Is there a way you could make the post reference my profile for source? Thanks!
@mkuegi@estherschindler Pretty impressive! And a great article as well!
Wouldn't have thought that a GA applied to some linear polynomials would be able to learn that.
I wonder if it is doomed to stagnate like the final models of the article, or if it would be able to further improve with more time.
Wow, I feel incredibly honored to be featured on this list! ๐
Thank you so much!โฅ๏ธ
I highly recommend anyone to check out ML-Agents if you're into machine learning! #mlagents
As we close out 2020, we'd like to highlight a few of our favorite #mlagents community projects ๐ค
Plus, check out how far we've come since Release 1 of ML-Agents and get a sneak peek into what we have planned for 2021 ๐
Find it all here: https://t.co/VlKzIaVmLA
Honored to receive the #SalzburgAG science award for my thesis! I was surprised by this lovely care package this morning.
Really made my day, thanks so much!
#wirsindzukunft
@mkuegi Seems like less than 2 weeks was a terrifyingly good estimate...
Then officials saw your tweet today and had to postpone the announcement to Saturday, so it is exactly 1 day > 2 weeks ;)
Starman, last seen leaving Earth, made its first close approach with Mars todayโwithin 0.05 astronomical units, or under 5 million miles, of the Red Planet
Today, after more than two and a half years of development and 15+ release updates, we are excited to announce that the ML-Agents Unity package has reached v1.0: https://t.co/H2D0aEgRiI @unity3d#AI#Simulation
After my last video a lot of people suggested to train an AI to parallel park.
So here it finally is: AI Learns to parallel park!
https://t.co/N1BgH7WKqr
Really enjoyed working with the latest @unity3d ML-Agents!
#madewithunity#MachineLearning#ReinforcementLearning#AI
I finally got to try out the newest @unity3d#MLAgents version after a long break. I really love how much progress has happened since my last project.
The workflow is so much more simple now and the training speed increased significantly!
@shihzy kudos to you and your team!
@phlhr@unity3d Thanks!
Doing that after training would probably cause the agent to be quite confused. Doing it during training could work quite well and would definitely help generalizability, but also increase the over all training time required.
Choose one out of three powerful Generals, assemble a mighty army and empower your General's strength to defeat your opponent on the continuously changing fields of exile!
Dromenon - Coming soon to https://t.co/PHKXD7PQln and Steam!
#madewithunity#gamedev#indiedev#game
@JanelleCShane @ElizaViera_ @unity3d Haha, pure chaos, I assume ๐
I would love to try to actually start the learning process again with them competing. Maybe even some adverserial rewards.
Just need to find some free time to do it ๐
@TshepoMoagi09 @unity3d Great observation! It was actually the exact same model. Someone on YT mentioned the same so I rewatched the entire footage. They ended up 25 : 20 (for green) but I only included the parts that were most fun to watch. Those parts happened to include many wins for the green car.