CtrlK
BlogDocsLog inGet started
Tessl Logo

tessl/pypi-rembg

Remove image background using advanced AI models including U-Net, BiRefNet, and SAM with support for multiple input formats and GPU acceleration

84

0.94x
Overview
Eval results
Files

rubric.jsonevals/scenario-6/

{
  "context": "This criteria evaluates how well the engineer uses rembg's alpha matting functionality for edge refinement. Alpha matting is an advanced technique in rembg that refines edges and handles semi-transparent regions by using pymatting algorithms with configurable thresholds and erosion parameters.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Uses rembg.remove",
      "description": "Uses the rembg.remove() function to perform background removal",
      "max_score": 20
    },
    {
      "name": "Enables alpha matting",
      "description": "Sets alpha_matting=True parameter to enable the alpha matting edge refinement feature",
      "max_score": 25
    },
    {
      "name": "Foreground threshold",
      "description": "Configures alpha_matting_foreground_threshold=240 to set the correct foreground detection threshold",
      "max_score": 15
    },
    {
      "name": "Background threshold",
      "description": "Configures alpha_matting_background_threshold=10 to set the correct background detection threshold",
      "max_score": 15
    },
    {
      "name": "Erosion size",
      "description": "Configures alpha_matting_erode_size=10 to set the correct erosion structure size for trimap generation",
      "max_score": 15
    },
    {
      "name": "Image I/O",
      "description": "Correctly reads the input image and writes the output image with transparency (PNG format)",
      "max_score": 10
    }
  ]
}

Install with Tessl CLI

npx tessl i tessl/pypi-rembg

tile.json