Referring Expression Comprehension on RefGTA (val)
71.66AccuracyReCLIP w/o relations
Evaluation Results
| Method | Links | |
|---|---|---|
| ReCLIP w/o relationsBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Ground-truth (GT)2022.04 | 71.66 | |
| ReCLIP w/o relationsBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Detected proposals filtered to only people (DT-P)2022.04 | 70.27 | |
| ReCLIPBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Ground-truth (GT)2022.04 | 69.84 | |
| ReCLIPBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Detected proposals filtered to only people (DT-P)2022.04 | 68.42 | |
| ReCLIP w/o relationsBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Detected proposals (DT)2022.04 | 60.98 | |
| ReCLIPBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Detected proposals (DT)2022.04 | 60.93 | |
| MDETRPretrained status=Not finetuned, Backbone=EfficientNet-B3, Proposals=None (End-to-End)2022.04 | 54.91 | |
| CLIP GradCAMBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Ground-truth (GT)2022.04 | 51.9 | |
| CLIP GradCAMBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Detected proposals filtered to only people (DT-P)2022.04 | 51.03 | |
| UNITER-LargeFinetuning Dataset=RefCOCOg, Proposals=Ground-truth (GT)2022.04 | 49.81 | |
| UNITER-LargeFinetuning Dataset=RefCOCO+, Proposals=Ground-truth (GT)2022.04 | 49.57 | |
| UNITER-LargeFinetuning Dataset=RefCOCOg, Proposals=Detected proposals filtered to only people (DT-P)2022.04 | 48.59 | |
| UNITER-LargeFinetuning Dataset=RefCOCO+, Proposals=Detected proposals filtered to only people (DT-P)2022.04 | 47.52 | |
| MDETRFinetuning Dataset=RefCOCO+, Backbone=EfficientNet-B3, Proposals=None (End-to-End)2022.04 | 38.49 | |
| MDETRFinetuning Dataset=RefCOCOg, Backbone=EfficientNet-B3, Proposals=None (End-to-End)2022.04 | 38.29 | |
| UNITER-LargeFinetuning Dataset=RefCOCO+, Proposals=Detected proposals (DT)2022.04 | 35.04 | |
| CLIP GradCAMBackbone=Ensemble of RN50x16 and ViT-B/32, Proposals=Detected proposals (DT)2022.04 | 33.66 | |
| UNITER-LargeFinetuning Dataset=RefCOCOg, Proposals=Detected proposals (DT)2022.04 | 27.58 | |
| RandomProposals=Ground-truth (GT)2022.04 | 27.03 | |
| RandomProposals=Detected proposals filtered to only people (DT-P)2022.04 | 21.53 | |
| RandomProposals=Detected proposals (DT)2022.04 | 4.86 |