Create AI portraits and short videos with strong likeness
With ArtCraft you can generate images and short video clips of yourself while keeping control over pose, lighting, and style. The workflow combines good references with structure-aware generation so the results look like you—and stay consistent across shots.
What you need
- A handful of clear photos that show your face from multiple angles, plus at least one full-body reference.
- Neutral lighting and no heavy filters. Avoid sunglasses or occlusions for the face.
Step-by-step
1. Prepare references: choose 4–8 images that cover front, 3/4, and profile views; include one where your outfit and background match the vibe you want. 2. Set the composition: on the 2D canvas, sketch layout and subject scale, or use the 3D scene to place camera and lighting. 3. Guide the model: add a concise prompt describing mood, setting, and attire; use masks to protect areas you don’t want to change. 4. Generate and refine: compare candidates, then fix details with inpainting (eyes, hairlines, hands, accessories). 5. Make it move: for short clips, render a small set of frames from slight camera or pose variations, then export as a looping video.
Likeness and safety tips
- Keep hair outline and eye region sharp in references; it improves identity consistency.
- Reuse the same camera distance and angle for a set; consistency beats variety for multi-shot galleries.
- Be mindful of where and how you publish. When sharing AI portraits, add context so viewers understand how they were made.