Guides

Video Background Removal

Core payload design and end-to-end processing flow.

Core Payload

{
  "video_url": "https://cdn.example.com/input.mp4",
  "output_format": "webm",
  "bg_type": "transparent",
  "model": "original"
}

Required Fields

FieldTypeRequiredDescription
video_urlstringyesSource video URL
modelstringyesoriginal | light | pro | human
bg_typestringyestransparent | color | image
output_formatstringyeswebm | mov | mp4 | gif | webp | png_sequence

Quality Controls

  • Prefer original or pro for complex accessories and thin edges.
  • Use human for portrait-oriented clips with frontal subjects.
  • For transparent output, use WebM in web apps and MOV/PNG in editing pipelines.

Quality Guardrails

  • Prefer representative source clips when tuning model defaults.
  • Keep input compression moderate to preserve edge details.
  • Use explicit output format policy for transparent vs non-transparent workflows.
  • Run A/B validation with the same clip set before changing defaults.

Implementation Checklist

  1. Define payload schema validation in backend before forwarding requests.
  2. Store model/output_format/background settings with each job record.
  3. Add internal quality review for difficult scenes (hair, glass, motion blur).
  4. Create runbook entries for model-specific failure cases.

Quality Evaluation Framework

MetricWhat to InspectTarget
Edge stabilityHair/cloth boundaries frame-to-frameNo obvious flicker
Foreground integrityAccessories and fine objects retainedMinimal detail loss
Temporal consistencyBrightness and alpha consistencyNo pulsing artifacts