Referring Image Segmentation on ReferIt (test)
76.18IoURefTr
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| RefTrVisual Backbone=RN101, Text Encoder=BERT-base2023.02 | 76.18 | — | — | — | — | — | — | — | — | |
| SHNetResolution=448x448, DenseCRF=false, Backbone=DeepLabv3+2021.04 | 69.19 | — | — | — | — | — | — | — | — | |
| SHNetResolution=320x320, DenseCRF=false, Backbone=DeepLabv3+2021.04 | 68.38 | — | — | — | — | — | — | — | — | |
| VIPAPublication=-, Vision Model=Swin-B, Language Model=BERT-B, Multi-dataset training=false2026.02 | 67.88 | — | — | — | — | — | 74.55 | — | — | |
| VIPAPublication=-, Vision Model=Swin-B, Language Model=BERT-B, Multi-dataset training=true2026.02 | 67.88 | — | — | — | — | — | 74.55 | — | — | |
| PolyFormer-LVisual Backbone=Swin-L, Text Encoder=BERT-base2023.02 | 67.22 | — | — | — | — | — | 72.6 | — | — | |
| ACMArchitecture=ACM2021.05 | 66.7 | — | — | — | — | — | — | — | — | |
| EFNDenseCRF=false2021.04 | 66.7 | — | — | — | — | — | — | — | — | |
| LSCMPost-processing=DenseCRF2021.05 | 66.57 | — | — | — | — | — | — | — | — | |
| LSCMDenseCRF=true2021.04 | 66.57 | — | — | — | — | — | — | — | — | |
| CGFormerPublication=CVPR 2023, Vision Model=Swin-B, Language Model=BERT-B, Multi-dataset training=false2026.02 | 66.42 | — | — | — | — | — | 73.36 | — | — | |
| VCMArchitecture=VCM2021.05 | 66.06 | — | — | — | — | — | — | — | — | |
| PolyFormer-BVisual Backbone=Swin-B, Text Encoder=BERT-base2023.02 | 65.98 | — | — | — | — | — | 71.91 | — | — | |
| PolyFormer-BPublication=CVPR 2023, Vision Model=Swin-B, Language Model=BERT-B, Multi-dataset training=true2026.02 | 65.98 | — | — | — | — | — | 71.91 | — | — | |
| TGFEfeature exchange rounds=32021.05 | 65.58 | — | — | — | — | — | — | — | — | |
| CMPCBackbone=DeepLab-101, Pretrained=PASCAL-VOC2020.10 | 65.53 | — | — | — | — | — | — | — | — | |
| CMPCPost-processing=DenseCRF2021.05 | 65.53 | — | — | — | — | — | — | — | — | |
| CMPCSource=CVPR 20202021.05 | 65.53 | — | — | — | — | — | — | — | — | |
| CMPCDenseCRF=true2021.04 | 65.53 | — | — | — | — | — | — | — | — | |
| STEP2020.10 | 64.13 | — | — | — | — | — | — | — | — | |
| STEP2021.05 | 64.13 | — | — | — | — | — | — | — | — | |
| STEP2021.05 | 64.13 | — | — | — | — | — | — | — | — | |
| STEPDenseCRF=true2021.04 | 64.13 | — | — | — | — | — | — | — | — | |
| CMSA NetworkSet=test2021.02 | 63.99 | 58.25 | 51.04 | 42.12 | 29.96 | 13.25 | — | — | — | |
| CMSA2019.04 | 63.8 | — | — | — | — | — | — | — | — | |
| CMSA2020.10 | 63.8 | — | — | — | — | — | — | — | — | |
| CMSAPost-processing=DenseCRF2021.05 | 63.8 | — | — | — | — | — | — | — | — | |
| CMSA2021.05 | 63.8 | — | — | — | — | — | — | — | — | |
| CMSADenseCRF=true2021.04 | 63.8 | — | — | — | — | — | — | — | — | |
| CMSASet=test2021.02 | 63.8 | 57.96 | 50.34 | 41.04 | 28.82 | 13.05 | — | — | — | |
| RRN2019.04 | 63.63 | — | — | — | — | — | — | — | — | |
| RRN2020.10 | 63.63 | — | — | — | — | — | — | — | — | |
| RRNPost-processing=DenseCRF2021.05 | 63.63 | — | — | — | — | — | — | — | — | |
| RRN2021.05 | 63.63 | — | — | — | — | — | — | — | — | |
| RRNDenseCRF=true2021.04 | 63.63 | — | — | — | — | — | — | — | — | |
| RRNSet=test2021.02 | 63.63 | 56.71 | 49.22 | 40.36 | 28.39 | 12.68 | — | — | — | |
| BRINetPost-processing=DenseCRF2021.05 | 63.46 | — | — | — | — | — | — | — | — | |
| BRINDenseCRF=true2021.04 | 63.46 | — | — | — | — | — | — | — | — | |
| ASGN2020.10 | 60.31 | — | — | — | — | — | — | — | — | |
| ASGN2021.05 | 60.31 | — | — | — | — | — | — | — | — | |
| KWA2021.05 | 59.19 | — | — | — | — | — | — | — | — | |
| KWA2019.04 | 59.09 | — | — | — | — | — | — | — | — | |
| KWA2020.10 | 59.09 | — | — | — | — | — | — | — | — | |
| KWA2021.05 | 59.09 | — | — | — | — | — | — | — | — | |
| KWASet=test2021.02 | 59.09 | 45.87 | 39.8 | 32.82 | 23.81 | 11.79 | — | — | — | |
| RMI2019.04 | 58.73 | — | — | — | — | — | — | — | — | |
| RMI2020.10 | 58.73 | — | — | — | — | — | — | — | — | |
| RMIPost-processing=DenseCRF2021.05 | 58.73 | — | — | — | — | — | — | — | — | |
| RMI2021.05 | 58.73 | — | — | — | — | — | — | — | — | |
| RMISet=test2021.02 | 58.73 | 46.08 | 38.9 | 30.77 | 20.62 | 8.54 | — | — | — | |
| DMN2019.04 | 52.81 | — | — | — | — | — | — | — | — | |
| DMN2020.10 | 52.81 | — | — | — | — | — | — | — | — | |
| DMN2021.05 | 52.81 | — | — | — | — | — | — | — | — | |
| DMN2021.05 | 52.81 | — | — | — | — | — | — | — | — | |
| DMNSet=test2021.02 | 52.81 | — | — | — | — | — | — | — | — | |
| LSTM-CNN2019.04 | 48.03 | — | — | — | — | — | — | — | — | |
| LSTM-CNN2020.10 | 48.03 | — | — | — | — | — | — | — | — | |
| LSTM-CNN2021.05 | 48.03 | — | — | — | — | — | — | — | — | |
| LSTM-CNN2021.05 | 48.03 | — | — | — | — | — | — | — | — | |
| BRINetVisual Backbone=RN101, Text Encoder=LSTM2023.02 | — | — | — | — | — | — | 63.11 | — | — | |
| CGFormerImage Encoder=Swin-B, Text Encoder=BERT2026.03 | — | — | — | — | — | — | — | 66.42 | 66.32 | |
| CMPCVisual Backbone=RN101, Text Encoder=LSTM2023.02 | — | — | — | — | — | — | 65.53 | — | — | |
| CMPC+Visual Backbone=RN101, Text Encoder=LSTM2023.02 | — | — | — | — | — | — | 65.58 | — | — | |
| EFNVisual Backbone=WRN101, Text Encoder=Bi-GRU2023.02 | — | — | — | — | — | — | 66.7 | — | — | |
| LSCMVisual Backbone=RN101, Text Encoder=LSTM2023.02 | — | — | — | — | — | — | 66.57 | — | — | |
| PolyFormerImage Encoder=Swin-B, Text Encoder=BERT, Trained on combined RefCOCO/+/G=true2026.03 | — | — | — | — | — | — | — | 65.98 | 71.91 | |
| ReSTRVisual Backbone=ViT-B, Text Encoder=Transformer2023.02 | — | — | — | — | — | — | 70.18 | — | — | |
| SSP-SAM-224Image Encoder=SAM-H, Text Encoder=CLIP, Input Resolution=2242026.03 | — | — | — | — | — | — | — | 67.26 | 72.65 | |
| SSP-SAM-336Image Encoder=SAM-H, Text Encoder=CLIP, Input Resolution=3362026.03 | — | — | — | — | — | — | — | 67.95 | 73.04 | |
| STEPVisual Backbone=RN101, Text Encoder=Bi-LSTM2023.02 | — | — | — | — | — | — | 64.13 | — | — |