embodystars

EmbodyStars – Visual Perception LoRA for Humanoid Robot Training

<Gallery />

Model description

EmbodyStars is a LoRA model designed to generate structured, realistic, and semantically consistent visual scenes for humanoid robot training and embodied AI systems.

The model focuses on improving: • Object recognition and spatial consistency • Human–environment interaction scenarios • Realistic indoor and task-oriented scenes • Visual diversity for synthetic training data

EmbodyStars is suitable for generating synthetic datasets used in perception training, simulation environments, and multimodal robot learning pipelines.

The model is intended to support vision-based learning for humanoid robots, reinforcement learning environments, and simulation-to-real transfer research.

Trigger words

You should use embodystars to trigger the image generation.

You should use robot vision to trigger the image generation.

You should use embodied scene to trigger the image generation.

Download model

Download them in the Files & versions tab.