Video Foreground Object Removal
Remove unwanted objects from video
Model Overview
Video Foreground Object Removal is a video cleanup model designed to remove foreground elements from footage using intelligent pixel reconstruction.
It combines a source video with a tracked matte to remove the selected object across frames, filling missing areas based on surrounding visual information to maintain continuity. Well suited for both creative editing and production workflows, it eliminates the need for a clean plate while preserving the original look of the footage.
Key Capabilities
Combines video and a tracked matte to remove foreground objects across frames without requiring a clean plate or paint.
Outputs background-reconstructed video and supports mask-based workflows for integration into pipelines.
Reconstructs removed areas with consistent clarity, color, and detail for natural results.
Pricing
Endpoint
Parameters
model
Model name:
inputFrameCount
Input video number of frames.
inputFrameRate
Input video frame rate in frames per second.
inputHeight
Input video height in pixels.
inputWidth
Input vidoe width in pixels.
outputHeight
Output video height in pixels.
outputWidth
Output video width in pixels.
videoBitDepth
Output video bit depth.
videoCodec
Output video codec.
Possible values: , ,
videoProfile
Output video profile.
Possible values: , ,
inputHeight
Input image height in pixels.
inputWidth
Input image width in pixels.
outputHeight
Output image height in pixels.
outputWidth
Output image width in pixels.
maskUri
URI for mask input.
referenceUri
URI for reference image input.
watermark
Whether to watermark the output.
Example Inputs
Last updated