DeepMind Unveils AI Models Empowering Robots to Perform Tasks Without Training
2025-03-13 / Read about 0 minute
Author:小编   

Google's DeepMind has introduced two groundbreaking AI models designed to elevate robots' capabilities in real-world applications. One of these models, dubbed Gemini Robotics, is a cutting-edge vision-language-action model that can comprehend new scenarios and execute a diverse array of tasks without the necessity for extensive training. This model significantly enhances robots' versatility, interactivity, and precision, allowing them to perform delicate operations like folding paper or removing bottle caps with ease. The second model focuses on spatial reasoning, empowering developers to run programs that leverage advanced reasoning abilities. These innovations represent a significant stride towards the development of more intelligent and versatile robotic technology.