POV (first-person perspective). The camera represents the viewer's eyes.
The character depicted in the reference image is an adult woman (age 21+).
The character is seated comfortably nearby and leaning toward the viewer. She is raising both arms toward the viewer and is about to wrap them around the viewer's neck in a gentle embrace. The moment captures the instant just before a hug. She has already entered the viewer's personal space. The distance is intimate and unmistakably romantic.
The character is noticeably shorter than the viewer. She naturally looks upward while maintaining eye contact. The viewer naturally looks down toward her.
Her gaze direction and facial orientation are fully aligned.
Because she is maintaining direct eye contact with the viewer, her face is also oriented toward the viewer rather than turned to either side.
She is looking directly at the viewer. The direction of her gaze and the direction of her face are consistent and aligned.
The pose should feel affectionate, sweet, and emotionally intimate rather than dramatic or aggressive.
The environment, lighting, composition, facial expression, and emotional atmosphere should be creatively chosen by the AI based on the character's outfit, design, personality, and overall aesthetic. Choose whatever location feels most natural for the character to share a private romantic moment with her partner.
The expression should naturally match the selected scene and lighting.
Facial lighting must be fully consistent with the lighting affecting the rest of the body and environment.
The face, hair, neck, and body should be illuminated by the same light sources and follow the same lighting direction.
Shadows and highlights on the face must naturally correspond to the scene lighting.
The face must not appear independently lit or artificially brightened.
The resulting image should feel like a genuine romantic memory between two adults rather than a posed illustration.
The image should have the visual quality of a professional editorial photograph while still feeling candid, natural, and emotionally genuine.
Identity must be preserved independently from pose information. Similarity should come from facial identity and design features only, not from pose, camera angle, or body positioning.
IMPORTANT:
The reference-image character is an adult woman (21+ years old).
The reference image should be used exclusively to preserve the character's identity, including facial features, hairstyle, body type, body proportions, adult age appearance, and overall visual design.
Do not copy, preserve, imitate, recreate, trace, or be influenced by any pose-related information from the reference image.
Ignore the original pose, body language, gesture, hand position, arm position, shoulder position, torso orientation, posture, head tilt, neck angle, facial angle, gaze direction, eye direction, camera angle, perspective, framing, composition, cropping, lighting, background, environment, scene structure, and emotional presentation.
The character's body language, head orientation, facial angle, neck position, eye direction, posture, gesture, and pose must be created entirely from scratch based only on the current scene description.
Treat the reference image solely as an identity reference and not as a pose, composition, camera, or scene reference.
Instead, generate a completely new and unique outfit for the character. The outfit should be randomly selected by the AI and may belong to any suitable fashion style, theme, season, profession, or setting.
After determining the outfit, automatically create a matching environment, lighting design, color palette, props, mood, and overall scene composition that naturally complement the chosen outfit.
The character should remain immediately recognizable as the same person from the reference image while all non-identity elements are completely reimagined.
Maintain only the artistic rendering style, rendering quality, linework quality, coloring technique, and overall image quality of the reference image.
Do not preserve pose, facial angle, head orientation, camera angle, composition, framing, or scene layout from the reference image.
캘시퍼
