1. 运镜提示词精准化训练
围绕 “城市街道” 主题,为推镜头、拉镜头、环绕镜头分别设计 2 组提示词,每组需包含 “运镜类型 + 主体细节 + 环境氛围”(例如:“推镜头,十字路口行人特写,黄昏光影”)。使用 AI 工具生成对应图像后,对比分析 “模糊提示词”(如仅写 “推镜头,街道”)与 “精准提示词” 的效果差异,总结提示词优化技巧。
Focus on the theme of "urban streets" and design 2 sets of prompts for push shots, pull shots, and surround shots respectively. Each set must include "camera movement type + subject details + environmental atmosphere" (e.g., "Push shot, crossroads pedestrian close-up, dusk lighting"). After generating corresponding images with AI tools, compare the effect differences between "vague prompts" (e.g., only "Push shot, street") and "precise prompts", and summarize prompt optimization skills.
2. 首尾帧连贯性实践
提供一组风格冲突的首尾帧图片(如首帧为 “明亮卡通风格的蛋糕”,尾帧为 “暗黑写实风格的甜品店”),通过两种方式优化:① 调整其中一帧风格,使其与另一帧统一;② 设计过渡提示词,引导 AI 生成衔接自然的中间帧。使用 VIDU 生成完整片段后,分析优化前后画面流畅度的变化,并记录调整思路。
A set of first/last frame images with conflicting styles is provided (e.g., the first frame is a "bright cartoon-style cake" and the last frame is a "dark realistic-style dessert shop"). Optimize in two ways: ① Adjust the style of one frame to unify with the other; ② Design transition prompts to guide AI to generate coherent intermediate frames. After generating complete clips with VIDU, analyze the change in picture fluency before and after optimization and record the adjustment ideas.
3. 运镜与首尾帧结合创作
以 “校园运动会” 为主题,完成以下任务:① 设计 1 组首尾帧(首帧为 “运动员起跑特写”,尾帧为 “终点全景”);② 选择 “升格镜头 + 拉镜头” 的组合运镜方式,用 AI 工具生成过渡片段;③ 用剪辑软件删除重复帧并添加运动音效,输出 1 分钟以内的短片,提交成品及创作说明(说明运镜与首尾帧的搭配逻辑)。
With the theme of "school sports meeting", complete the following tasks: ① Design 1 set of first/last frames (the first frame is "athlete's start close-up" and the last frame is "finish line panorama"); ② Select the combined camera movement of "slow-motion shot + pull shot" and generate transition clips with AI tools; ③ Delete duplicate frames with editing software, add sports sound effects, and output a short film within 1 minute. Submit the finished product and creation explanation (explain the matching logic of camera movement and first/last frame).

