A naukri.com initiative
The Robot Report
3w
293
Google DeepMind introduces two Gemini-based models to bring AI to the real world
- Google's robotics team, with expertise in machine learning, engineering, and physics simulation, aims to tackle challenges in AI-powered robots.
- DeepMind introduces Gemini Robotics and Gemini Robotics-ER models based on Gemini 2.0 for real-world applications.
- Gemini Robotics is an advanced vision-language-action model designed to control robots directly through physical actions.
- Gemini Robotics-ER enhances spatial understanding, enabling a wider range of tasks for robots.
- DeepMind partners with Apptronik to build humanoid robots using Gemini 2.0.
- Gemini Robotics focuses on generality, interactivity, and dexterity crucial for real-world robotics applications.
- Gemini Robotics can adapt to novel situations, comprehend natural language instructions, and adjust actions dynamically.
- Gemini Robotics' dexterity allows it to tackle complex tasks like origami folding and adapt to various robot form factors.
- Gemini Robotics-ER emphasizes spatial reasoning and enables connections with existing low-level controllers.
- DeepMind prioritizes safety in robotics by integrating safety features in Gemini Robotics-ER for collision avoidance and dynamic stability.
Read Full Article
17 Likes
For uninterrupted reading, download the app