Jacob Perrin’s Post

I have been part of the program committee for the Symposium on Machine Learning for Autonomous Driving (2023). 📑 The papers of 2023 have been great to review. I like the notion and sense of urgency that today drives research towards making these systems save and trustworthy - can't wait to see which work makes it into the workshop. 🗓 The workshop will be held on 14th Dec, 2023 co-located with NeurIPS 2023) for more information upfront visit: https://ml4ad.github.io/ 👭 Peer-reviewing is a lot of work and lives at the core of scientific progress. At the same time it is rewarding, as it allows to dive deep into the current methods and technology. So next time you are asked to review something, don't think twice and do it! 🚌 Thanks for the organization Rowan McAllister (Waymo) - and thanks for having me for the 4th consecutive year now, I appreciate it. #autonomousdriving #artificialintelligence #Science #ADsafety #DeepLearning #SaveAI #ZFambassador ZF Group

When #AI enables machines to see, understand, and integrate into the physical world through vision like humans do, what does this mean for autonomous driving? How can we imagine the future of intelligence? Join us on July 16th for the #Nullmax 2024 Tech Conference to find out. #Nullmax #AI #AIGC #AutonomousDriving

Motivations of Using Deep Reinforcement Learning: An Autonomous Vehicle Case Study DRL can be used in financial services, marketing, healthcare, robotics, computer vision, and autonomous vehicles (AVs). In this article, we elaborate on the motivations for using DRL in AVs. This article is a valuable reference for PhD students and researchers interested in applying DRL in different scenarios particularly AVs, providing a comprehensive overview of the motivations and benefits of DRL. 1. DRL Can Optimize Large State and Action Spaces 2. DRL is Low-Cost in Implementation and Inference 3. DRL Has a Long-Term Goal 4. DRL is Adaptive 5. DRL Can Be Model-Free 6. DRL Works with Outdated Data 7. DRL Can Reduce Signaling Overhead in AVs 8. DRL Scales to Optimize Many Factors Jointly Read further at the following link: https://lnkd.in/dbXctQtR #DeepReinforcementLearning #AutonomousVehicles #MachineLearning #AIResearch #UAVTechnology #AVOptimization #Robotics

"Did you know that the term 'data science' was coined as recently as 2001? It's incredible to see how this field has evolved and grown exponentially since then. From personalized medicine to autonomous vehicles, the potential of machine learning and AI in transforming our world is truly inspiring. Let's continue to explore the possibilities and drive innovation in this exciting industry. #DataScience #MachineLearning #ArtificialIntelligence #Innovation #FutureTech"

Is it time to say goodbye to the automated driving dream, or can we still turn the tide? At TNO, we believe in combining the strengths of both learning and symbolic #AI for the next generation of autonomous driving systems. Curious? Then this is your last chance to join our free online session on the 11th of July and discover how AI can be a solution and a challenge. > https://tno.to/ayk Jan-Pieter Paardekooper (TNO) Riccardo Donà (European Commission JRC) Henk Goossens Olaf Op den Camp Erwin de Gelder Arturo Tejada Ruiz #TNO #AV #AutomatedDriving

🚀 Excited to share our latest blog post on "Cohere3D: Exploiting Temporal Coherence for Unsupervised Representation Learning of Vision-based Autonomous Driving." Our team introduces a novel contrastive learning algorithm, Cohere3D, to learn coherent instance representations in long-term input sequences, facilitating instance-level correspondence across multiple input frames in autonomous driving scenarios. The approach shows promising results in improving data efficiency and task performance across various downstream perception, prediction, and planning tasks. Dive into the details and explore the potential of Cohere3D at https://bit.ly/3wxkcCo. Let's shape the future of autonomous driving together! 🚗🔍 #autonomousdriving #deeplearning #representationlearning

#didyouknow The first artificial intelligence model, known as the Logic Theorist, was developed in 1955 by Allen Newell, J.C. Shaw, and Herbert A. Simon. This groundbreaking program was capable of proving mathematical theorems and is considered one of the earliest examples of machine learning. Since then, AI has evolved dramatically, revolutionizing industries and shaping the way we interact with technology today. From virtual assistants to autonomous vehicles, the possibilities are endless with AI! #DidYouKnow #AIHistory #TechnologyEvolution 🚀🔍

Last week, I had the pleasure to present our paper „Can You Handle The Truth?: Investigating the Effects of Uncertainty of Deep Learning Models on Users of Autonomous Vehicles“ at the 22nd IEEE International Symposium on Mixed and Augmented Reality #ISMAR, that was held in Sydney. In this paper, we show the importance of uncertainty quantification as a central part of the output of the model, especially in safety-critical applications, such as #autonomous_driving. We consider 3 main aspects that are critical to the safety of the vehicle: Situation uncertainty, waypoint uncertainty, and trajectory uncertainty. While quantifying the uncertainty of deep learning models can seem challenging in some set-ups, model-agnostic frameworks, such as conformal prediction can make those accessible with validity guarantees.

In the fascinating world of Artificial Intelligence, every moment is an opportunity for discovery. Did you know that the first AI program was written in 1951? Since then, AI has transcended sci-fi to become a potent driver of real-world innovation - from enhancing medical diagnostics to pioneering autonomous vehicles. Let's celebrate the perseverance that continues to transform dreams into tangible technology. Remember, every line of code brings us closer to the future. #Innovation #DataScience #ArtificialIntelligence #MachineLearning #AIInspiration #TechFuture

SEE YOU IN HAWAII! Call for papers to the 1st WACV Workshop on Large Language and Vision Models for Autonomous Driving (LLVM-AD) during IEEE/CVF Winter Conference on Applications of Computer Vision - WACV 2024. This workshop aims to bring together professionals from academia and industry to explore the application of large language and vision models in autonomous driving. As part of this initiative, two open-source datasets are released for real-world traffic language understanding. Workshop topics include: ·      Vision-Language Models in Autonomous Systems ·      Large Language Models for Autonomous Driving Applications ·      Vision Models for Autonomous Driving ·      Semantics and Scene Understanding in Autonomous Driving ·      Human-Vehicle Interaction ·      Domain Adaptation and Transfer Learning in Autonomous Driving ·      Safety in Autonomous Driving ·      Interpretable AI in Autonomous Driving Submission Deadline: October 23, 2023. Workshop Website: https://llvm-ad.github.io. #llms #computervision #autonomousdriving #autonomousvehicles #humancomputerinteraction #hawaii