本发明公开了一种图片生成方法、显示方法、设备、存储介质及程序产品。该方法基于输入的图像或多模态信息，通过特征提取、风格控制与语义生成模块，生成目标图像并在显示设备上进行渲染。本发明能够在提高图像生成效率与质量的同时，支持个性化生成与多维语义控制，适用于智能视觉系统、虚拟现实、广告创作及艺术生成等多种应用场景。
English Abstract:
The invention discloses an image generation method, display method, device, storage medium, and program product. The method extracts and integrates multimodal features to generate target images with controllable styles and semantics. It enhances generation efficiency and image quality while supporting personalized and multi-dimensional visual generation, applicable to intelligent vision systems, virtual reality, advertising, and artistic creation.