How do you generate a driving scene that stays consistent over time, across cameras, and under controlled changes? These are exactly the types of problems we solve.
Mobileyeโs Visual GenAI team develops image and video generation capabilities for autonomous driving, with a growing focus on video and world models.
Weโre looking for a hands-on researcher to help shape our research direction and turn ideas into working capabilities, with access to large proprietary driving datasets and close collaboration with the teams developing Mobileyeโs next generation of autonomous driving technology.
Mobileye changes the way we drive, from preventing accidents to semi and fully autonomous vehicles. If you are an excellent, bright, hands-on person with a passion to make a difference come to lead the revolution!
What will your job look like:
Research and develop image and video generation models, with a focus on temporal consistency, multiview coherence, and controllable driving-scene generation.
Explore and adapt state-of-the-art world models for autonomous driving, helping shape the teamโs research and technical direction.
Own the full development lifecycle, from research and experimentation to model training, evaluation, and integration into real-world workflows.
Build scalable, clean Python and PyTorch implementations for training and evaluating generative models.
Collaborate with AI researchers and perception teams to develop generative capabilities for training and evaluating autonomous driving systems.
All you need is:
M.Sc. or Ph.D. in Computer Science, Electrical Engineering, or a related field.
3+ years of industry experience in Deep Learning and Computer Vision, preferably working on generative AI or foundation models.
A hands-on approach and the ability to turn research ideas into working solutions.
Nice to have:
Hands-on experience training image or video generation models, including diffusion and flow-based models.
Experience with world models, multiview generation, 3D scene representations, or temporal modeling.
Experience training or adapting foundation models, including Vision-Language Models (VLMs) and other multimodal models.
Experience with distributed training and large-scale datasets.
Publications in top-tier conferences such as CVPR, NeurIPS, ICCV, or ECCV.
