withoutBG Open-Weight Model vs IS-Net (isnet-general-use)
37 examples, 37 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) | 0.07063 ✓ | 0.08337 |
| MAE (Overall Accuracy) | 0.06432 ✓ | 0.08659 |
| Connectivity | 0.06406 ✓ | 0.08562 |
- MGE
- 0.11187✓
- MAE
- 0.02393✓
- Conn.
- 0.02366✓
- MGE
- 0.12449
- MAE
- 0.07297
- Conn.
- 0.06980
- MGE
- 0.07491
- MAE
- 0.20595✓
- Conn.
- 0.20605✓
- MGE
- 0.06655✓
- MAE
- 0.22381
- Conn.
- 0.22360
- MGE
- 0.07847
- MAE
- 0.01629✓
- Conn.
- 0.01623✓
- MGE
- 0.07657✓
- MAE
- 0.06137
- Conn.
- 0.06394
- MGE
- 0.01863✓
- MAE
- 0.00476✓
- Conn.
- 0.00413✓
- MGE
- 0.02615
- MAE
- 0.01341
- Conn.
- 0.01185
- MGE
- 0.04704
- MAE
- 0.00841
- Conn.
- 0.00826
- MGE
- 0.03013✓
- MAE
- 0.00528✓
- Conn.
- 0.00464✓
- MGE
- 0.04385✓
- MAE
- 0.01141✓
- Conn.
- 0.01000✓
- MGE
- 0.06105
- MAE
- 0.02702
- Conn.
- 0.02405
- MGE
- 0.03629
- MAE
- 0.01250✓
- Conn.
- 0.01173✓
- MGE
- 0.03522✓
- MAE
- 0.02263
- Conn.
- 0.02122
- MGE
- 0.12069
- MAE
- 0.01999
- Conn.
- 0.01970
- MGE
- 0.08820✓
- MAE
- 0.01739✓
- Conn.
- 0.01604✓
- MGE
- 0.04500
- MAE
- 0.00749✓
- Conn.
- 0.00713✓
- MGE
- 0.04268✓
- MAE
- 0.06100
- Conn.
- 0.05949
- MGE
- 0.02821✓
- MAE
- 0.00834✓
- Conn.
- 0.00806✓
- MGE
- 0.05411
- MAE
- 0.02930
- Conn.
- 0.02823
- MGE
- 0.06872
- MAE
- 0.01673✓
- Conn.
- 0.01649✓
- MGE
- 0.06718✓
- MAE
- 0.01971
- Conn.
- 0.01778
- MGE
- 0.06195
- MAE
- 0.01177✓
- Conn.
- 0.01189
- MGE
- 0.05507✓
- MAE
- 0.01320
- Conn.
- 0.01178✓
- MGE
- 0.02073
- MAE
- 0.00482✓
- Conn.
- 0.00464✓
- MGE
- 0.01957✓
- MAE
- 0.00658
- Conn.
- 0.00563
- MGE
- 0.06311
- MAE
- 0.01859
- Conn.
- 0.01844
- MGE
- 0.04879✓
- MAE
- 0.01407✓
- Conn.
- 0.01423✓
- MGE
- 0.11979✓
- MAE
- 0.02627✓
- Conn.
- 0.02612✓
- MGE
- 0.12660
- MAE
- 0.08300
- Conn.
- 0.08018
- MGE
- 0.02311✓
- MAE
- 0.00796✓
- Conn.
- 0.00769✓
- MGE
- 0.03633
- MAE
- 0.02418
- Conn.
- 0.02425
- MGE
- 0.01507✓
- MAE
- 0.00650✓
- Conn.
- 0.00573✓
- MGE
- 0.03501
- MAE
- 0.02718
- Conn.
- 0.02617
- MGE
- 0.09111✓
- MAE
- 0.01771✓
- Conn.
- 0.01754✓
- MGE
- 0.17796
- MAE
- 0.27290
- Conn.
- 0.27664
- MGE
- 0.05274✓
- MAE
- 0.02840✓
- Conn.
- 0.02826✓
- MGE
- 0.08141
- MAE
- 0.05662
- Conn.
- 0.05412
- MGE
- 0.37456
- MAE
- 0.49888
- Conn.
- 0.49455
- MGE
- 0.33448✓
- MAE
- 0.46974✓
- Conn.
- 0.48246✓
- MGE
- 0.05988
- MAE
- 0.01390✓
- Conn.
- 0.01386✓
- MGE
- 0.05076✓
- MAE
- 0.01701
- Conn.
- 0.01449
- MGE
- 0.01330✓
- MAE
- 0.00272✓
- Conn.
- 0.00254✓
- MGE
- 0.01823
- MAE
- 0.01416
- Conn.
- 0.01331
- MGE
- 0.02282✓
- MAE
- 0.00565✓
- Conn.
- 0.00489✓
- MGE
- 0.03478
- MAE
- 0.02152
- Conn.
- 0.01752
- MGE
- 0.07366
- MAE
- 0.01213
- Conn.
- 0.01216
- MGE
- 0.06092✓
- MAE
- 0.01037✓
- Conn.
- 0.00923✓
- MGE
- 0.03789
- MAE
- 0.02922✓
- Conn.
- 0.03011✓
- MGE
- 0.03269✓
- MAE
- 0.03657
- Conn.
- 0.03447
- MGE
- 0.02475✓
- MAE
- 0.00395✓
- Conn.
- 0.00373✓
- MGE
- 0.02644
- MAE
- 0.00537
- Conn.
- 0.00451
- MGE
- 0.01271✓
- MAE
- 0.00301✓
- Conn.
- 0.00277✓
- MGE
- 0.01871
- MAE
- 0.00612
- Conn.
- 0.00494
- MGE
- 0.02503✓
- MAE
- 0.00752✓
- Conn.
- 0.00688✓
- MGE
- 0.03361
- MAE
- 0.02629
- Conn.
- 0.02443
- MGE
- 0.03017✓
- MAE
- 0.00736✓
- Conn.
- 0.00695✓
- MGE
- 0.06968
- MAE
- 0.02905
- Conn.
- 0.02494
- MGE
- 0.04226✓
- MAE
- 0.01824✓
- Conn.
- 0.01716✓
- MGE
- 0.08788
- MAE
- 0.05431
- Conn.
- 0.05123
- MGE
- 0.28577✓
- MAE
- 0.93281
- Conn.
- 0.93738
- MGE
- 0.56277
- MAE
- 0.90155✓
- Conn.
- 0.90653✓
- MGE
- 0.11198
- MAE
- 0.02944
- Conn.
- 0.02949
- MGE
- 0.10144✓
- MAE
- 0.02800✓
- Conn.
- 0.02763✓
- MGE
- 0.17475
- MAE
- 0.02962
- Conn.
- 0.02905
- MGE
- 0.11987✓
- MAE
- 0.02251✓
- Conn.
- 0.02072✓
- MGE
- 0.11440
- MAE
- 0.29795✓
- Conn.
- 0.29837✓
- MGE
- 0.08196✓
- MAE
- 0.35117
- Conn.
- 0.35466
- MGE
- 0.03113✓
- MAE
- 0.00814✓
- Conn.
- 0.00763✓
- MGE
- 0.06492
- MAE
- 0.02734
- Conn.
- 0.02722
- MGE
- 0.03992✓
- MAE
- 0.01483✓
- Conn.
- 0.01451✓
- MGE
- 0.10094
- MAE
- 0.10320
- Conn.
- 0.09046
- MGE
- 0.01719✓
- MAE
- 0.00674✓
- Conn.
- 0.00642✓
- 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.