Oxford Cognitive Robotics Group

The Cognitive Robotics Group investigates problems in perception, inference, learning, and interaction to enable robots to operate in open-ended environments.

[Oxford Cognitive Robotics Group @CognitiveRobots]

Last week, the CRG celebrated our supervisor Lars Kunze @kula78 as he started a new position as a full professor with @BristolRobotLab.

Lars will continue leading the CRG as a visiting fellow, but please join us in celebrating his new role! 🎉 Congrats, Lars!

[Oxford Cognitive Robotics Group @CognitiveRobots]

Last week, Lars @kula78 attended the
@RealAAAI Symposium on AI Trustworthiness and Risk Assessment for Challenged Contexts (ATRACC) where he presented our CARLA Responsible AI Autonomous Driving Challenge. 🚗

Learn more & participate: https://t.co/eUWqup1thX

#CARLARAIChallenge

[Oxford Cognitive Robotics Group @CognitiveRobots]

It was exciting to engage with the community and see strong interest in the project.

In the coming weeks, you can expect to see our custom-designed backpack making its way around Oxford routes! 🚴

[Oxford Cognitive Robotics Group @CognitiveRobots]

Today, CRG and @MRG_Oxford were on Broad Street demonstrating the ongoing RobotCycle project which is measuring cyclist safety around Oxford!

More info: https://t.co/koOjkEOtOp

Project leads:
@panefimia
@d_d_martini

[Efimia Panagiotaki @panefimia]

Can neural networks reason like pointcloud registration algorithms?
Excited to share our latest paper, "NAR-*ICP: Neural Execution of Classical ICP-based Pointcloud Registration Algorithms", with @PetarV_93, @d_d_martini, and @kula78!
https://t.co/DuEwZvbvLb
More info below 🤖

[Lars Kunze @kula78]

🚗 Exciting News! 🚗
We’re thrilled to announce the CARLA RAI Autonomous Driving Challenge!
Evaluate driving agents against #responsibleAI (#RAI) indicators: https://t.co/5lcMJeikVz
Paper: https://t.co/OsVXDknFs1
@tas_hub @awscloud @carlasimulator @eval_ai
@SteadyBits @i_m_prs

[Oxford Cognitive Robotics Group @CognitiveRobots]

Last Friday we attended the RoboTips showcase event as a culmination of several years' effort towards responsible robotics development. Congrats @Marenka @alan_winfield @kula78 and team! Thanks for the demo @oxfordrobots

Check out the project here: https://t.co/bOuplpPVwt

[Lars Kunze @kula78]

Had a fantastic time at the #RoboTIPS event. Was great to see so many people from academia and industry as well as policymakers engaged in #ResponsibleRobotics. Congrats to @Marenka @alan_winfield and team! And huge thanks to @oxfordrobots for a demo on #causal #explainability!

[Lars Kunze @kula78]

Listening to @MichaelGroom16 giving a great talk on #safety for mobile manipulation for domestic robots at the @Dyson Institute of Engineering and Technology (https://t.co/wQVhR1G97e) at the AI and the Future of Engineering Symposium 2024.

[Oxford Cognitive Robotics Group @CognitiveRobots]

Excellent talk by Dr. Jo-Ann Pattinson (@ITSLeeds) about legal implications of AI-generated evidence for motor vehicle accident disputes

Last week, research fellows on our funded RAILS project attended the Trustworthy Autonomous System (TAS-R) workshop at the University of Leeds

[Oxford Cognitive Robotics Group @CognitiveRobots]

🚗 Today at IEEE IV 2024, Samuel Sze presented his paper, "Real-time 3D semantic occupancy prediction for autonomous vehicles using memory-efficient sparse convolution" 👏

📚 Check it out here: https://t.co/nIpSGWo9mi

#IV2024 #IEEE #semanticsegmentation

[Oxford Cognitive Robotics Group @CognitiveRobots]

Stay tuned for further updates from the Cognitive Robotics Group this week at IEEE IV 2024! 👋

[Oxford Cognitive Robotics Group @CognitiveRobots]

📢 Live from Jeju Island, South Korea:

Two of our students, Benjamin Hardin @b_b_hardin and Samuel Sze are at the 2024 IEEE Intelligent Vehicles Symposium this week to present their work!

At today's sessions they presented two papers👇🧵

#IV2024 #remotedriving #cyclistsafety

[Oxford Cognitive Robotics Group @CognitiveRobots]

Samuel (https://t.co/peYvcPlHr7) presented the RobotCycle poster on behalf of @panefimia at the Socially Interactive Autonomous Mobility Workshop.

The work used a novel sensing unit and traffic modelling approach to create a dataset of cycling safety.

https://t.co/aEtOGIaDLm

[Daniel Omeiza @SteadyBits]

New RSS paper! RAG-Driver, a #MLLM that leverages in-context learning and #RAG techniques to predict vehicle control signals and generate explanations, achieving SOTA generalisability on the BDD-X set and on our new driving commentary dataset. #XAI #AutonomousVehicles #LLM

[Arto Bendiken @bendiken]

Catching @kula78's engrossing talk Making Robots Explainable and Trustworthy at the Actionable Knowledge Representation and Reasoning for Robots (AKR³) workshop #ESWC2024

[Oxford Robotics Institute @oxfordrobots]

Thanks @BristolRobotLab for coming to visit yesterday! We loved all the questions and discussions. It was a great chance to share highlights from our research and demonstrate some of the new robots about the lab!

[Oxford Cognitive Robotics Group @CognitiveRobots]

Greetings from #ICRA2024 🇯🇵 🤖 👋

@RCannizzaro01 @panefimia Divya Thuremella, & @kula78

[Oxford Cognitive Robotics Group @CognitiveRobots]

At #ICRA2024? @panefimia will be at the Fields Robotics workshop presenting a poster about RobotCycle: assessing the safety of Oxford cyclists using a novel sensing unit and traffic modelling approach.

Come say hi between 9am to 1pm tomorrow (13/5)! 👋

https://t.co/aEtOGIa5VO

[Oxford Cognitive Robotics Group @CognitiveRobots]

A massive effort led by @panefimia, @kula78, and our team collaborating with many others across @oxfordrobots and @oxengsci! A huge amount of time and energy has been poured into developing this 👏 Great work, everyone! #ICRA2024

[Lars Kunze @kula78]

On my way to #ICRA2024!! This year, we will be sharing insights on our #offroad #radar dataset (https://t.co/n4m9lflUZZ), #RobotCycle (https://t.co/xnfkzVLB0p), and risk-aware trajectory prediction. See you in #Japan!

@oxfordrobots @CognitiveRobots @BristolRobotLab