Jensen's inequality

cseweb.ucsd.edu/~tzli/novelt...
I gave an internal talk at UCSD last year regarding "novelty" in computer science research. In it I "debunked" some of the myth people seem to have about what is good research in computer science these days. People seemed to like it, so I thought I should share.

Our new E2E driving method, TransFuser v6, is out on ArXiv.
It outperforms all other methods on CARLA by a wide margin, 95 DS on Bench2Drive!
We show that minimizing the asymmetry between data annotator and policy is key for strong IL results.

Code, models, and paper:
ln2697.github.io/lead/

Why study self-driving? Isn't it "solved"?

What's left to do in self-driving given Waymo is taking off? An argument that it's still a great research problem:
open.substack.com/pub/emergere...

Attending #Neurips2025? Get your personalized Scholar Inbox conference program now to easily navigate the poster sessions and find what you are looking for:
www.scholar-inbox.com/conference/n...

Scholar Inbox for NeurIPS is live now.

Launching the Physical AI AV Dataset! 🚀

huggingface.co/datasets/nvi...

One of the largest, most diverse & commercially usable open-source datasets for AVs.
- 1727 hours of driving data
- Camera, LiDAR, & radar
- 25 countries, 2500+ cities

This is just the beginning, more features to come!

Takeru Miyato - AI Research Scientist PhD student at University of Tübingen working on artificial intelligence, machine learning, and deep learning research.

Congratulations to our PhD student @takerumiyato.bsky.social for winning the Google PhD Fellowship in the category "Machine Learning and ML Foundations". Takeru is pioneering new neural architectures that improve generalization and efficiency. Check out his research: takerum.github.io

🚗 Join us for the RealADSim Workshop @ ICCV 2025!
🗓 Oct 19 morning | 📍 Hawaii Conv. Center, 305A
Featuring exciting talks and winner presentations from our challenges (70+ regs, ~30 subs). Don’t miss it! 🎉
#ICCV2025

ELLIS PhD Program: Call for Applications 2025 The ELLIS mission is to create a diverse European network that promotes research excellence and advances breakthroughs in AI, as well as a pan-European PhD program to educate the next generation of AI...

ellis.eu/news/ellis-p...

Announcing the @iccv.bsky.social NAVSIM Challenge! What's new? We're testing not only on real recordings, but also perturbed futures generated from the real ones via pseudo-simulation! $8K in prizes + several $1.5k travel grants. Submit by September 20! opendrivelab.com/challenge2025/ 🧵👇

Christian will be mind-blown, once he finds out about Corsairs VENGEANCE RGB DDR5 Light Enhancement Kit. Still mind blowing to me, that this product exists.

The RGB RAM sticks give at least 10-20 FPS more.

NeurIPS PCs: Borderline Reject: "Please use sparingly"
NeurIPS reviewers: BR go BR BR BR BR BR BR

One concern that I have as an AI researcher when publishing code is that it can potentially be used in dual-use applications.
To solve this, we propose Civil Software Licenses. They prevent dual-use while being minimal in the restrictions they impose:

civil-software-licenses.github.io

New YouTube playlist category is Relaxing self-driving agents to vibe to

GitHub - autonomousvision/CaRL: [ArXiv 2025] CaRL: Learning Scalable Planning Policies with Simple Rewards [ArXiv 2025] CaRL: Learning Scalable Planning Policies with Simple Rewards - autonomousvision/CaRL

We have released the code for our work, CaRL: Learning Scalable Planning Policies with Simple Rewards.

The repository contains the first public code base for training RL agents with the CARLA leaderboard 2.0 and nuPlan.

github.com/autonomousvi...

Back in 2023, I worked on rule-based planning for the nuPlan challenge and was skeptical that learned methods could do well on this benchmark. Turns out, RL is the answer. The CaRL code is out now. Also check out this 2h+ demo video of our policy: youtu.be/_godUKkICec

CaRL: Learning Scalable Planning Policies with Simple Rewards YouTube video by Daniel Dauner

In case you find it as relaxing as we do: Here is a 2h+ video of our autonomous RL driving agent CaRL in action! @danieldauner.bsky.social @bernhard-jaeger.bsky.social @kashyap7x.bsky.social
youtube.com/watch?v=_god...

Christina's new work PrITTI generates high-quality, controllable 3D semantic urban scenes in a compact primitive-based representation using a latent diffusion model. This enables applications like scene editing, inpainting, and photo-realistic street view synthesis!
raniatze.github.io/pritti/

The 2025 Waymo Challenge results are out! We achieved second place in the End-to-End Driving Challenge with DiffusionLTF and third place in the Scenario Generation Challenge with SHRED. Our approaches prioritized simplicity, and we're preparing to make them publicly available soon.

Is it? Sometimes it's the unexpected results that reveal something more interesting.

🚗 Pseudo-simulation combines the efficiency of open-loop and robustness of closed-loop evaluation. It uses real data + 3D Gaussian Splatting synthetic views to assess error recovery, achieving strong correlation with closed-loop simulations while requiring 6x less compute. arxiv.org/abs/2506.04218

EMPERROR is a new generative perception error model for probing self-driving planners. It imitates modern 3D detectors, maintains temporal consistency, and reveals brittleness in IL-based planning - raising collision rates by up to 85%! #RA_L #robotics #AI - lasnik.github.io/emperror/

📣 Excited to share our #CVPR2025 Spotlight paper and my internship project @wayve: SimLingo.
A Vision-Language-Action (VLA) model that achieves state-of-the-art driving performance with language capabilities.

Code: github.com/RenzKa/simli...
Paper: arxiv.org/abs/2503.09594

I've always found the default Alt+Tab behavior in Ubuntu frustrating. Recently, I discovered this plugin for GNOME distros. Personally, it’s a game changer and enables switching windows specific to the application, monitor, and workspaces. Highly recommend!
github.com/G-dH/advance...

📢 New paper CVPR 25!
Can meshes capture fuzzy geometry? Volumetric Surfaces uses adaptive textured shells to model hair, fur without the splatting / volume overhead. It’s fast, looks great, and runs in real time even on budget phones.
🔗 autonomousvision.github.io/volsurfs/
📄 arxiv.org/pdf/2409.02482

How do you know someone's a Bayesian? Don't worry, they'll tell you.

Waited to finish reading it but I wanted to highlight the nice reward design here, which alleviates many of the failures of reward shaping

Introducing CaRL: Learning Scalable Planning Policies with Simple Rewards
We show how simple rewards enable scaling up PPO for planning.
CaRL outperforms all prior learning-based approaches on nuPlan Val14 and CARLA longest6 v2, using less inference compute.
arxiv.org/abs/2504.17838