We’re launching Lyria 3.5 in Google Flow Music, with advances across musicality, lyrics, vocals, and creative control
deepmind.google·6d ago
AI Engineering
DeepMind Blog·Apr 13·5 min read
Gemini Robotics-ER 1.6: Powering real-world robotics tasks through enhanced embodied reasoning
TL;DR
Robots need to reason about their physical environments to be effective, rather than just following instructions. Gemini Robotics-ER 1.6 enhances embodied reasoning, allowing robots to understand spatial contexts and perform tasks autonomously.
✦ Why It Matters
Engineers can leverage Gemini Robotics-ER 1.6 to develop more capable and autonomous robotic systems for diverse applications.
Key Takeaways
- Explore Gemini Robotics-ER 1.6 for improved task execution.
- Utilize enhanced spatial reasoning for complex robotics tasks.
- Integrate instrument reading capabilities into your applications.
- Leverage safety improvements for better compliance in robotics.
How It Works
Gemini Robotics-ER 1.6 employs agentic vision, which combines visual reasoning with code execution. This allows the model to zoom into images for better detail, use pointing to identify relevant features, and execute code to derive accurate readings from instruments.
By integrating these capabilities, the model can interpret complex visual inputs and make informed decisions based on its understanding of the physical world.