id=1 sec=5 prompt= subject_definitions: <Subject 1> is the… | Deepfake AI视频

生成此视频所用的文字转视频提示词

id=1 sec=5 prompt= subject_definitions: <Subject 1> is the girl from the input image — she is the speaker (S1), whose voice comes from <Audio 1>. Positioned LEFT, recording a selfie, cheerful expression. <Subject 2> is the street BEHIND her — a busy city sidewalk with storefronts, pedestrians, and parked cars. summary: [reference generation] The target video opens with <Subject 1> recording a selfie on the street, speaking to the camera. <Subject 2> is visible in the background. retention_analysis: <Subject 1> (appears in [Shot 1]): fully_preserved - the girl is positioned left, recording a selfie, cheerful expression. <Subject 2> (appears in [Shot 1]): fully_preserved - the street with storefronts, pedestrians, and parked cars is retained. detailed_description: [Shot 1] Clean modern video, natural daylight, sharp image, smooth handheld selfie shot. <Subject 1> is on the LEFT, recording a selfie on the street, holding the camera at arm's length, framing her face and shoulders. Behind her: <Subject 2> — a busy city sidewalk with storefronts, a few pedestrians walking past in soft motion blur, and parked cars along the curb. She looks straight into the lens, turns her head slowly to the right, then back to the left, then to center, showing her face from both sides, and breaks into a bright smile. (S1) says cheerfully: <d>[English] Hi, I'm model number thirty!</d> FINAL: She holds the smile at center, tilts her head slightly, and keeps her eyes on the lens until the clip ends. overall_soundscape: Distant traffic hum, pedestrian chatter, car horns, footsteps on pavement – clean and natural, not compressed. non_diegetic_music: N/A