Dime Dime Baby AI 视频生成器

派对舞步、咬面包和凝视

Create a new photorealistic short video. Image 1 is the identity reference for the main adult, not a first frame. Preserve the person's recognizable face, hair, skin tone, build and clothing throughout. A warm Mexican-style party hall with colorful paper decorations and soft string lights. A few adult guests remain out of focus in the distant background. The reference adult is the only main performer, wearing their own outfit from Image 1 and holding a pink concha sweet bread in one hand. A fixed eye-level camera, continuous take. 0-2.5 seconds: the performer dance-walks toward the camera with small bent-knee steps and a relaxed shoulder sway, framed head to knees. 2.5-4 seconds: they stop near the camera and take one small bite of the concha, then lower it so their face is fully visible. 4-6 seconds: hold a funny, completely deadpan stare directly into the lens while gently chewing. 6-9 seconds: turn around and dance-walk away, swaying shoulders back toward the party. Preserve the approach, bread bite, stare and retreat in that order. Audio: newly composed light Latin-style percussion and quiet party ambience, instrumental only. Natural hands and anatomy; no face morphing, text, logos or watermark. Original scene and audio only; no film footage, existing recordings, songs or voice imitation. Required settings override optional notes: same reference adult, the stated setting and action sequence, newly composed instrumental audio, vertical 9:16, 9 seconds, 480P.