New week, new opportunity! 🚀Become a Research Scientist in Data & AI Model Auditing (f/m/div.)✨. If you would like to develop novel methods to identify systematic errors, publish results & apply them to real-world problems, we welcome your application ▶bit.ly/45EW9Nv

We are conducting cutting-edge #AI research! We focus on methods & technologies like foundation models and generative AI. 🚀These and more AI methods are used in various Bosch sectors & applications, resulting in innovative products & business models 🤩➡ bit.ly/3RySPAc

The countdown is on: Tomorrow is the last chance to apply to the #ELLISPhD Program this year! Join our network and kick-start your career in modern #AI research in Europe! ellis.eu/news/ellis-phd…

Can multilingual training boost #NLP in low-resource languages? Check out GradSim, a gradient similarity-based method for language grouping, setting new sentiment analysis benchmarks.📚 Discover insights on model layers and language traits. 📃➡ bit.ly/3snAZWN

It is international Student Day! 👩‍🎓 You are a student & would like to contribute your ideas not only in lectures but also in corporate practice? Take a look at the many different opportunities to help us at BCAI shape the future while gaining experience▶bit.ly/3XVeMJk

Ready for a new adventure? 🚀 Shape the future of cyber-physical systems that will be increasingly distributed across a device-edge-cloud continuum. Apply as Research Engineer for Software Engineering for Distributed Cyber-Pysical Systems (f/m/div.) ➡bit.ly/46OECne

📢Interested in Object-Centric Slot Representation w/o predefined slots for a better scene understanding? 🚀Thrilled to share our #BMVC23 paper "Enhancing Interpretable Object Abstraction via Clustering-based Slot Initialization" ✨ 🌐gaobaoding.github.io/publication/bm…

📅 Save the date! #NeurIPS2023 will take place from Dec 10 to Dec 16. Our #Bosch team will be on-site in New Orleans, USA 📍. Don't miss the opportunity to network an get in contact with our #AI experts. 🚀Find more information on #NeurIPS2023 here ➡ bit.ly/3gQTcGh

How can robots adapt better to unseen situations or new requirements? This research offers a solution to optimize Gaussian mixture policies via optimal transport & Riemannian optimization, for various scenarios, from reaching motions to multi-goal tasks ➡️ bit.ly/46BdJTu

Make sure to check out the paper “Towards Anytime Classification in Early-Exit Architectures by Enforcing Conditional Monotonicity” accepted at #NeurIPS2023 to learn more how an early-exit network can become gradually confident 📑➡ bit.ly/3Ql3zjE

Are you looking for a role where your skills can shape the future of autonomous driving? 🌐 We're actively seeking a dynamic Research Scientist in Neural Rendering for Multi-Modal Synthetic Data Generation (f/m/div.) to join our forward-thinking team! 💡▶ bit.ly/44afiHz

Curious how recent #BayesianDeepLearning methods perform on data beyond the training distribution? At #NeurIPS2023, there's large-scale survey on the #WILDS dataset, exploring fine-tuning, transformers, and deep ensembles in real-world scenarios. 📃➡️ bit.ly/3PXR25r

A new opportunity awaits you! Become #Research Engineer - Tools and Methods for Embedded #AI System Design (f/m/div.) at #Bosch! You will optimize and deploy neural networks for different hardware targets.✨ We are looking forward meeting you! Apply Now ➡ bit.ly/3Gie5DL

Dive into the intricacies of pseudo-likelihood inference, and discover how it combines a likelihood kernel, a neural density estimator, and an integral probability metric to infer a distribution over parameters. Check out the paper accepted @NeurIPS23 📑➡ bit.ly/3QqWtKr

Let's get it started! ✨ On December 10, the #NeurIPS2023 conference will start. Visit us at our booth and make sure to check our conference schedule beforehand! Our #Bosch #AI experts are already looking forward meeting you in New Orleans, USA 🙌

Ever considered optimizing model performance on a tight annotation budget? In data-efficient #DL, #BCAI researchers (US) explore data pre-selection, endorsing a joint vision-text feature space. UP-DP, an innovative approach, attains up to 20% gain. 🌐➡ vimeo.com/892506139?shar…

Discover #GradOrth: an innovative solution from the #BCAI team in Sunnyvale, presented at #NeurIPS23. This approach prevents incorrect predictions on significantly different data, enhancing OOD detection with minimal code integration. 🌟 Learn more ➡ vimeo.com/892507946?shar…

Can prompt learning overcome vision-language model challenges? The study "Text-driven Prompt Generation for Vision-Language Models in Federated Learning" delves into this query, aiming to boost CLIP-based models' adaptability in federated learning. 🌐➡ bit.ly/3QWLn16

When academia and industry unite, remarkable things happen! 🤝 Meet Stefan Kurz, Chief Expert for #HybridModeling at #Bosch, embodying the best of both worlds. We're thrilled to announce his new additional role as Professor of Practice at ETH Zurich, starting Feb 2024! 🚀

👇Check out our work presented at NeurIPs now! Early-exit networks aim at adjusting the computation effort according to the actual complexity of classifying each sample. We further encourage them to become gradually confident. More processing only for better prediction quality!