withoutBG Open-Weight Model vs IS-Net (isnet-general-use)
37 examples, 36 with ground truth. Lower is better. Use the view controls on each row to switch between transparency, chroma key, and alpha matte on all outputs.
| Metric | withoutBG Open-Weight Model | IS-Net |
|---|---|---|
| MGE (Edge Quality) | 0.03887 ✓ | 0.07006 |
| MAE (Overall Accuracy) | 0.03281 ✓ | 0.06395 |
| Connectivity | 0.03238 ✓ | 0.06281 |
- MGE
- 0.03827✓
- MAE
- 0.00801✓
- Conn.
- 0.00745✓
- MGE
- 0.12449
- MAE
- 0.07297
- Conn.
- 0.06980
- MGE
- 0.02810✓
- MAE
- 0.02909✓
- Conn.
- 0.02904✓
- MGE
- 0.06655
- MAE
- 0.22381
- Conn.
- 0.22360
- MGE
- 0.03931✓
- MAE
- 0.01038✓
- Conn.
- 0.01018✓
- MGE
- 0.07657
- MAE
- 0.06137
- Conn.
- 0.06394
- MGE
- 0.01556✓
- MAE
- 0.00499✓
- Conn.
- 0.00447✓
- MGE
- 0.02615
- MAE
- 0.01341
- Conn.
- 0.01185
- MGE
- 0.01735✓
- MAE
- 0.00687
- Conn.
- 0.00668
- MGE
- 0.03013
- MAE
- 0.00528✓
- Conn.
- 0.00464✓
- MGE
- 0.03203✓
- MAE
- 0.01223✓
- Conn.
- 0.01087✓
- MGE
- 0.06105
- MAE
- 0.02702
- Conn.
- 0.02405
- MGE
- 0.01712✓
- MAE
- 0.01079✓
- Conn.
- 0.01024✓
- MGE
- 0.03522
- MAE
- 0.02263
- Conn.
- 0.02122
- MGE
- 0.06369✓
- MAE
- 0.01273✓
- Conn.
- 0.01265✓
- MGE
- 0.08820
- MAE
- 0.01739
- Conn.
- 0.01604
- MGE
- 0.02636✓
- MAE
- 0.00503✓
- Conn.
- 0.00479✓
- MGE
- 0.04268
- MAE
- 0.06100
- Conn.
- 0.05949
- MGE
- 0.01869✓
- MAE
- 0.00816✓
- Conn.
- 0.00786✓
- MGE
- 0.05411
- MAE
- 0.02930
- Conn.
- 0.02823
- MGE
- 0.02853✓
- MAE
- 0.00936✓
- Conn.
- 0.00923✓
- MGE
- 0.06718
- MAE
- 0.01971
- Conn.
- 0.01778
- MGE
- 0.04587✓
- MAE
- 0.01240✓
- Conn.
- 0.01258
- MGE
- 0.05507
- MAE
- 0.01320
- Conn.
- 0.01178✓
- MGE
- 0.01224✓
- MAE
- 0.00520✓
- Conn.
- 0.00515✓
- MGE
- 0.01957
- MAE
- 0.00658
- Conn.
- 0.00563
- MGE
- 0.03009✓
- MAE
- 0.01084✓
- Conn.
- 0.01069✓
- MGE
- 0.04879
- MAE
- 0.01407
- Conn.
- 0.01423
- MGE
- 0.04795✓
- MAE
- 0.01123✓
- Conn.
- 0.01103✓
- MGE
- 0.12660
- MAE
- 0.08300
- Conn.
- 0.08018
- MGE
- 0.01477✓
- MAE
- 0.00731✓
- Conn.
- 0.00697✓
- MGE
- 0.03633
- MAE
- 0.02418
- Conn.
- 0.02425
- MGE
- 0.01251✓
- MAE
- 0.00568✓
- Conn.
- 0.00480✓
- MGE
- 0.03501
- MAE
- 0.02718
- Conn.
- 0.02617
- MGE
- 0.05039✓
- MAE
- 0.02393✓
- Conn.
- 0.02378✓
- MGE
- 0.17796
- MAE
- 0.27290
- Conn.
- 0.27664
- MGE
- 0.03115✓
- MAE
- 0.01899✓
- Conn.
- 0.01896✓
- MGE
- 0.08141
- MAE
- 0.05662
- Conn.
- 0.05412
- MGE
- 0.37414
- MAE
- 0.49631
- Conn.
- 0.49210
- MGE
- 0.33448✓
- MAE
- 0.46974✓
- Conn.
- 0.48246✓
- MGE
- 0.02245✓
- MAE
- 0.00523✓
- Conn.
- 0.00503✓
- MGE
- 0.05076
- MAE
- 0.01701
- Conn.
- 0.01449
- MGE
- 0.00932✓
- MAE
- 0.00731✓
- Conn.
- 0.00723✓
- MGE
- 0.01823
- MAE
- 0.01416
- Conn.
- 0.01331
- MGE
- 0.01754✓
- MAE
- 0.00608✓
- Conn.
- 0.00528✓
- MGE
- 0.03478
- MAE
- 0.02152
- Conn.
- 0.01752
- MGE
- 0.03535✓
- MAE
- 0.00615✓
- Conn.
- 0.00610✓
- MGE
- 0.06092
- MAE
- 0.01037
- Conn.
- 0.00923
- MGE
- 0.01507✓
- MAE
- 0.02355✓
- Conn.
- 0.02418✓
- MGE
- 0.03269
- MAE
- 0.03657
- Conn.
- 0.03447
- MGE
- 0.01807✓
- MAE
- 0.00420✓
- Conn.
- 0.00401✓
- MGE
- 0.02644
- MAE
- 0.00537
- Conn.
- 0.00451
- MGE
- 0.01089✓
- MAE
- 0.00275✓
- Conn.
- 0.00248✓
- MGE
- 0.01871
- MAE
- 0.00612
- Conn.
- 0.00494
- MGE
- 0.01419✓
- MAE
- 0.00609✓
- Conn.
- 0.00537✓
- MGE
- 0.03361
- MAE
- 0.02629
- Conn.
- 0.02443
- MGE
- 0.02140✓
- MAE
- 0.00739✓
- Conn.
- 0.00718✓
- MGE
- 0.06968
- MAE
- 0.02905
- Conn.
- 0.02494
- MGE
- 0.03300✓
- MAE
- 0.01786✓
- Conn.
- 0.01614✓
- MGE
- 0.08788
- MAE
- 0.05431
- Conn.
- 0.05123
- MGE
- ?
- MAE
- ?
- Conn.
- ?
- MGE
- 0.56277
- MAE
- 0.90155
- Conn.
- 0.90653
- MGE
- 0.06131✓
- MAE
- 0.09372
- Conn.
- 0.09310
- MGE
- 0.10144
- MAE
- 0.02800✓
- Conn.
- 0.02763✓
- MGE
- 0.09468✓
- MAE
- 0.01882✓
- Conn.
- 0.01849✓
- MGE
- 0.11987
- MAE
- 0.02251
- Conn.
- 0.02072
- MGE
- 0.04716✓
- MAE
- 0.24497✓
- Conn.
- 0.24520✓
- MGE
- 0.08196
- MAE
- 0.35117
- Conn.
- 0.35466
- MGE
- 0.01980✓
- MAE
- 0.00724✓
- Conn.
- 0.00659✓
- MGE
- 0.06492
- MAE
- 0.02734
- Conn.
- 0.02722
- MGE
- 0.02499✓
- MAE
- 0.01318✓
- Conn.
- 0.01269✓
- MGE
- 0.10094
- MAE
- 0.10320
- Conn.
- 0.09046
- MGE
- 0.01003✓
- MAE
- 0.00726✓
- Conn.
- 0.00712✓
- MGE
- 0.03163
- MAE
- 0.02776
- Conn.
- 0.02542
Methodology
- Alpha mattes are grayscale masks encoding per-pixel foreground probability. They capture fine edge detail (hair, fur, semi-transparent regions) that binary masks discard.
- Metrics are computed by comparing predicted mattes against a ground-truth reference: MGE measures edge sharpness, MAE measures pixel-level accuracy, and Connectivity penalises fragmented foreground regions. All three are lower-is-better.
- Scores are shown as ? when no ground-truth matte is available for that image. See the alpha matting evaluation benchmark for details on the evaluation methodology.