Video-to-video restyles the whole frame. A region edit does not. You draw a box (or a rough mask), you say what should happen inside it, and the model is supposed to leave everything outside alone: the camera move, the room, the other person. That is the point of the tool — a surgical pass on a take you already like, not a second generate of the shot.
Video inpainting is the underlying operation when the request is 'remove this' or 'replace this with X.' Segmentation is how a precise mask gets tracked when a box is too crude. A region edit is the product name for that loop on a timeline: pick the span, pick the region, prompt the change, preview the join. The failure mode is bleed — the box was tight and the new pixels have nowhere to seam, or the box was loose and a shoulder you wanted kept got rewritten.
Identity and physics still apply inside the box. A hand you asked the model to invent will hover. A label you asked it to letter will shimmer if temporal consistency is weak. Prefer a real product plate in the region, or a still you own, over a prompt that has to spell the logo. Then watch the edges at full speed, not paused on the prettiest frame.
In practice
- Box generously around the object, then check that nothing you need to keep is inside the box.
- Prompt the replacement, not the thing you are removing.
- If the camera is moving, watch the box track before you trust the fill.
The mistake to avoid
Region-editing a change that contradicts the source motion — a new camera, a new pose — and blaming the mask. The region is a patch, not a new shot.
Related terms
Video inpainting
Video inpainting regenerates a masked region through time, so a removal or replacement holds across frames instead of flickering per still.
Segmentation
Segmentation labels which pixels in each frame belong to a chosen object, producing a mask that other tools then act on.
Video-to-video
Video-to-video takes finished footage in and returns altered footage, using the original clip as the structural reference for every frame.
Inpainting
Inpainting regenerates a region you have masked while leaving the rest of the picture untouched, so a change stays local.
Text-to-video
Text-to-video is generation from a written prompt alone — you describe a shot, the model invents every frame of it, and no image or footage goes in.
The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync, in your browser or on your phone.
Free on the web and on iPhone, one account.