AI Видео

Use the person in the first image as a precise model: keep their face, body, and hairstyle 100% intact. Do not alter or reimagine any aspect of their appearance. Take the person's face and figure from the first image, and the pose, emotions, and background from the second. The pose must be identical—no deviations, no adjustments, no stylization. Replace the person in the second image with only the person in the first. Maintain a natural facial expression that matches the emotions in the second image. The result should look like a real photo taken on an iPhone—realistic, lifelike, not artificial.

Take the attached video as a reference for movement and transfer its dynamics to the character in the photo, maintaining its identity: face, body proportions, facial features, skin texture, hair, clothing fit, and overall silhouette, without any changes or stylistic deviations. The movement should be reproduced as accurately as possible, frame by frame, including rhythm, tempo, accelerations and decelerations, shifts in the center of gravity, foot placement (including rolling), balance, arm trajectories, finger movements, head rotations, gaze direction, micro-expressions, and breathing. The physics of motion must remain realistic: take into account inertia, muscle tension, joint limitations, and interactions with surfaces. Any artifacts, such as sliding legs, stretching limbs, shaking, leaps, or unnatural transitions, must be excluded. No additional movements, effects, speed changes, or visual techniques can be added, only those present in the original video. The result must be clean and static.