Instance Segmentation on NC4K
52.9APOVCIS (task-specific)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| OVCIS (task-specific)Learning Paradigm=closed-set supervised learning, Fine-tuning=true, Params (Millions)=28.7/1522.72023.12 | 52.9 | 76.8 | 55.9 | |
| DCNetLearning Paradigm=closed-set supervised learning, Params (Millions)=53.4/53.42023.12 | 52.8 | 77.1 | 56.5 | |
| UQFormerLearning Paradigm=closed-set supervised learning, Params (Millions)=37.5/37.52023.12 | 47.2 | 74.2 | 49.2 | |
| Mask2FormerLearning Paradigm=closed-set supervised learning, Params (Millions)=43.9/43.92023.12 | 45.8 | 73.6 | 47.5 | |
| CamoFALearning Paradigm=closed-set supervised learning2023.12 | 45 | 75.7 | 44.3 | |
| MaskFormerLearning Paradigm=closed-set supervised learning, Params (Millions)=45.0/45.02023.12 | 44.6 | 71.9 | 45.8 | |
| OSFormerLearning Paradigm=closed-set supervised learning, Params (Millions)=46.6/46.62023.12 | 42.5 | 72.5 | 42.3 | |
| MSPNetLearning Paradigm=closed-set supervised learning, Params (Millions)=48.1/48.12023.12 | 41.8 | 71.8 | 42.3 | |
| SOLOv2Learning Paradigm=closed-set supervised learning, Params (Millions)=46.2/46.22023.12 | 34.4 | 65.9 | 31.9 | |
| CondInstLearning Paradigm=closed-set supervised learning, Params (Millions)=34.1/34.12023.12 | 33.4 | 67.4 | 29.4 | |
| QueryInstLearning Paradigm=closed-set supervised learning, Params (Millions)=172.5/172.52023.12 | 33 | 66.7 | 29.4 | |
| YOLACTLearning Paradigm=closed-set supervised learning, Params (Millions)=35.3/35.32023.12 | 32.1 | 65.3 | 27.9 | |
| MS R-CNNLearning Paradigm=closed-set supervised learning, Params (Millions)=60.0/60.62023.12 | 31 | 58.7 | 29.4 | |
| HTCLearning Paradigm=closed-set supervised learning, Params (Millions)=76.9/76.92023.12 | 29.8 | 59 | 26.6 | |
| Cascade R-CNNLearning Paradigm=closed-set supervised learning, Params (Millions)=71.7/71.72023.12 | 29.5 | 60.8 | 24.8 | |
| Mask TransfinerLearning Paradigm=closed-set supervised learning, Params (Millions)=44.3/44.32023.12 | 29.4 | 56.7 | 27.2 | |
| SOTRLearning Paradigm=closed-set supervised learning, Params (Millions)=63.1/63.12023.12 | 29.3 | 61 | 25.6 | |
| Mask R-CNNLearning Paradigm=closed-set supervised learning, Params (Millions)=43.9/43.92023.12 | 27.7 | 58.6 | 22.7 | |
| BlendMaskLearning Paradigm=closed-set supervised learning, Params (Millions)=35.8/35.82023.12 | 27.7 | 56.7 | 24.2 | |
| OVCISLearning Paradigm=open-vocab T2I, Fine-tuning=false, Params (Millions)=28.7/1522.72023.12 | 24.8 | 44.2 | 23.9 | |
| ODISELearning Paradigm=open-vocab T2I, Fine-tuning=false, Params (Millions)=28.1/1522.12023.12 | 22.9 | 37.2 | 21.4 | |
| TPNetLearning Paradigm=open-vocab VLM, Fine-tuning=false, Params (Millions)=71.78/71.782023.12 | 21.4 | 48.3 | 16.6 | |
| OpenSeeDLearning Paradigm=open-vocab VLM, Fine-tuning=false, Params (Millions)=116.2/116.22023.12 | 9.3 | 14.5 | 9.8 | |
| SEEMLearning Paradigm=open-vocab VLM, Fine-tuning=false, Params (Millions)=415.3/415.32023.12 | 9.2 | 12.7 | 9.9 | |
| MasQCLIPLearning Paradigm=open-vocab VLM, Fine-tuning=false, Params (Millions)=375.2/357.22023.12 | 8 | 7.6 | 8.4 | |
| MaskCLIPLearning Paradigm=open-vocab VLM, Fine-tuning=false, Params (Millions)=542.0/542.02023.12 | 6.3 | 5.6 | 6.5 | |
| X-DecoderLearning Paradigm=open-vocab VLM, Fine-tuning=false, Params (Millions)=38.3/38.32023.12 | 3.9 | 8.1 | 3.4 |