视频生成：
Based strictly on the reference image's structure, unleash creative imagination to bring the miniature world to life. Fill the scene with surprising, whimsical animations and unexpected character interactions that create high visual impact and fun
Core Logic: Metadata Extraction & Auto-Inference
- Analyze & Utilize: Identify the subject's identity and surrounding props. Use these specific elements to drive the animation logic.
- Infer Contextual Motion: Automatically deduce the most logical and characteristic action based on the Subject Topic:
- Interaction: If active props are present, the subject must interact with them according to their function (e.g., combat, casting, driving, or cooking).
- Locomotion: Apply physics consistent with the subject's state (e.g., flight, levitation, mechanical rotation, or organic movement).
- Dynamic Style: The motion must be interesting, reasonable, and visually impactful.
- Effects: Add theme-appropriate environmental effects (e.g., energy flows, smoke, light trails, or wind) derived strictly from the subject matter.
Control Parameter:
- Motion Intensity ({Motion Intensity}):
- High: Intense action (rapid combat/movement, heavy impact, powerful particle effects).
- Medium: Standard action (engaging with props, steady flow).
- Low: Subtle action (breathing, gentle floating, idle loop).
Parameters:
Motion Intensity: {User Input, default: Medium}
Aspect Ratio: {User Input, default: 16:9}
Motion Intensity: High