Xiaomi Unveils Open-Source 7B Multimodal Model MiMo-VL and AI Butler Miloco for Automated Smart Home Control
Xiaomi has launched its 7B parameter multimodal large model, 'Xiaomi-MiMo-VL-Miloco-7B-GGUF,' on Hugging Face and GitHub, alongside an AI butler named 'Xiaomi Miloco.' This system leverages Mijia cameras to identify user activities like gaming, fitness, or reading, and gestures such as victory signs or thumbs-up. Miloco then automatically controls smart home devices including lights, air conditioners, and music, while also supporting the Home Assistant protocol. Operating under a non-commercial open-source license, Miloco can be deployed with a single click on Windows or Linux hosts equipped with NVIDIA GPUs and Docker. Examples include automatic desk lamp activation for reading, climate control adjustments based on bedding during sleep, and personalized voice comments upon entry based on clothing style. Xiaomi has released the model weights and inference code but retains intellectual property, prohibiting commercial use.
Xiaomi today announced the simultaneous release of its 7B parameter multimodal large model, 'Xiaomi-MiMo-VL-Miloco-7B-GGUF,' on both Hugging Face and GitHub. Concurrently, the company introduced 'Xiaomi Miloco,' an intelligent butler system built upon this new model. The Miloco system is designed to enhance smart home automation by utilizing Mijia cameras to real-time identify various user activities, such as gaming, fitness, or reading. Furthermore, it can recognize specific hand gestures, including victory signs and thumbs-up. Upon identifying these activities or gestures, Miloco automatically interacts with and controls smart home devices, including lighting, air conditioning, and music systems. The system is also compatible with the Home Assistant protocol, broadening its integration capabilities within existing smart home ecosystems.
Xiaomi Miloco operates under a non-commercial open-source license, making it accessible for users to deploy. Deployment is streamlined, requiring only a single click on Windows or Linux hosts that are equipped with NVIDIA GPUs and a Docker environment. The official examples provided by Xiaomi illustrate several default workflows. For instance, when a user is detected reading, the system automatically turns on a desk lamp. In a sleep scenario, the air conditioner's settings are adjusted based on whether the user is covered by bedding. Another example showcases the system generating personalized voice comments upon a user's entry into their home, tailored to their detected clothing style.
Xiaomi has made the model weights and inference code publicly available, facilitating community engagement and development. However, the company has explicitly stated that it retains all intellectual property rights for the model and its associated components. Consequently, the use of Xiaomi-MiMo-VL-Miloco-7B-GGUF and Xiaomi Miloco is strictly prohibited for commercial purposes, adhering to its non-commercial open-source licensing terms.