HUGE AI Video Platform Home
0:00
0:00
Download
AI Video from prompt: Use the uploaded image as the first frame and as

Sora 2
Use the uploaded image as the first frame and as the appearance reference. Generate a photorealistic vertical video 9:16, duration 6 seconds, 24–30 fps. Scene: a fair/light-skinned man lying on a pillow in a room, soft evening/home lighting, calm atmosphere, cinematic color grading, natural skin texture, realistic facial animation. Camera movement: smooth camera pull-back (zoom out / dolly out) from an extreme close-up of the face to a wider shot (head + part of shoulders/chest and the pillow visible). Slight depth of field, no shake. Character animation (very natural): 0–2 s: the man looks sleepy, steady breathing, subtle micro-movements in the face. 2–4.5 s: slowly closes his eyes, eyelids lower smoothly, expression relaxes. 4.5–6 s: after falling asleep, his head gently turns slightly to the left side (not abrupt), stays relaxed, calm breathing. Constraints / negative prompt: no change of identity or facial features, no “plastic” skin, no distorted eyes/mouth/teeth, no morphing, no extra limbs, no flicker, no sudden exposure jumps, no camera jitter, no artifacts on the pillow or background, no change of clothing or environment.

huge