Remove Objects and Repair Images with AI

Video Weaver offers two image tools for cleanup. Remove objects from the frame marks an area to remove; Advanced AI fill paints a mask and generates replacement content for a still image. Both work from a selected image clip in the timeline.

Remove an object with a selection box

1. Select an image clip and expand Remove objects from the frame in the selected clip's tools panel.
2. Drag a rectangle around the unwanted object or watermark in the preview. Use Clear selection if you need to redraw it.
3. Select Remove selected area and wait for the processed preview.
4. Choose Replace selected image to use the result in place of the selected clip, or Add edited image to timeline to keep the original too.

The first use prepares about 28.1 MB of object-removal data. The rectangle marks the same position throughout a video segment, so this is intended for still images or objects that stay in one place. For a video, the original audio is kept.

Fill a painted area with generative AI

1. Select an image clip and expand Advanced AI fill.
2. Paint over the object or damaged region in the preview. Adjust Brush for broad or precise marks; use Erase mask or Clear mask to correct the selection.
3. Select Fill marked area. This model does not take a text prompt, so the surrounding image guides the fill.
4. Review the repaired preview, then replace the selected image clip or add the result as a separate timeline clip.

Advanced AI fill runs on-device at 512 × 512. Its first use downloads about 1.27 GB, requires WebGPU, and limits the output's long edge to 512 pixels. It does not fall back to CPU/WASM. Use it for small image repairs when your device has WebGPU and enough storage; use the object-removal box for a smaller, simpler selection.