Scene Recognition on SUN RGB-D Scene (test)
60.7Acc (RGB-D)Finetuned ResNet101-RNN
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Finetuned ResNet101-RNNBackbone=ResNet101, Recurrent component=RNN, Finetuning=true2020.04 | 60.7 | 58.5 | 50.1 | — | — | |
| TRecgNet Aug - ResNet101Backbone=ResNet1012020.04 | 59.8 | 54.2 | 49.3 | — | — | |
| CBCL2020.04 | 59.5 | 48.8 | 37.3 | — | — | |
| 2D-3D FusionNet2020.04 | 58.6 | 56.4 | 44.1 | — | — | |
| TRecgNet Aug - ResNet18Backbone=ResNet182020.04 | 58.5 | 53.8 | 49.3 | — | — | |
| ASK2020.04 | 57.3 | — | — | — | — | |
| TRecgNet Aug2020.04 | 56.7 | 50.6 | 47.9 | — | — | |
| MAPNet2020.04 | 56.2 | — | — | — | — | |
| MSN2020.04 | 56.2 | — | — | — | — | |
| G-L-SOOR2020.04 | 55.5 | 50.5 | 44.1 | — | — | |
| Cross-Modal Graph2020.04 | 55.1 | 45.7 | — | — | — | |
| DF2Net2020.04 | 54.6 | — | — | — | — | |
| RGB-D-OB2020.04 | 53.8 | 42.4 | — | — | — | |
| Fix ResNet101-RNNBackbone=ResNet101, Recurrent component=RNN, Finetuning=false2020.04 | 53.1 | 50.8 | 38.6 | — | — | |
| RGB-D-CNN2020.04 | 52.4 | 42.7 | 42.4 | — | — | |
| MSMM2020.04 | 52.3 | 41.5 | 40.1 | — | — | |
| LM-CNN2020.04 | 48.7 | 44.3 | 34.6 | — | — | |
| Places CNN-RCNN2020.04 | 48.1 | 40.4 | 36.3 | — | — | |
| MDSI-CNN2020.04 | 45.2 | 39.6 | 35.2 | — | — | |
| HP-CNN-T2020.04 | 42.2 | 38.8 | 28.5 | — | — | |
| RAGC2020.04 | 42.1 | — | — | — | — | |
| DMFF2020.04 | 41.5 | 37 | — | — | — | |
| SS-CNN-R62020.04 | 41.3 | 36.1 | — | — | — | |
| Places CNN-RBF SVMClassifier=RBF SVM2020.04 | 39 | 38.1 | 27.7 | — | — | |
| Places CNN-Lin SVMClassifier=Linear SVM2020.04 | 37.2 | 35.6 | 25.5 | — | — | |
| CLIP²3D Representation=Point Cloud2026.01 | — | — | — | — | 41.39 | |
| EMSAFormerBackbone=SwinV2-T-128, Augmentation Strategy=Multi-Aug2023.06 | — | — | — | 64.96 | — | |
| EMSAFormerBackbone=SwinV2-T-128, Augmentation Strategy=Multi-Aug, Decoder Configuration=Sem(SegFormer)2023.06 | — | — | — | 64.5 | — | |
| EMSANetBackbone=2x ResNet1012023.06 | — | — | — | 61.21 | — | |
| EMSANetBackbone=2x ResNet34-NBt1D2023.06 | — | — | — | 62.66 | — | |
| GiT-BTraining=multi-task, Specific Modules=None, Examples Num=1, #Params=131M2024.03 | — | — | — | — | 30.9 | |
| GiT-BTraining=universal, Specific Modules=None, Examples Num=1, #Params=131M2024.03 | — | — | — | — | 37.5 | |
| GiT-HTraining=universal, Specific Modules=None, Examples Num=1, #Params=756M2024.03 | — | — | — | — | 42.5 | |
| GiT-LTraining=universal, Specific Modules=None, Examples Num=1, #Params=387M2024.03 | — | — | — | — | 39.9 | |
| PanopticNDT (EMSANet-R34-NBt1D)Protocol=Application network, LR=0.00052023.09 | — | — | — | 58.56 | — | |
| PanopticNDT (EMSANet-R34-NBt1D)Protocol=Fine-tuning, LR=0.0022023.09 | — | — | — | 60.17 | — | |
| ReCo+Protocol=Zero-Shot Transfer, Specific Modules=DeiT-SIN, Examples Num=4, #Params=46M2024.03 | — | — | — | — | 24.2 | |
| TIGAUSSIAN3D Representation=3DGS2026.01 | — | — | — | — | 76.46 | |
| Token FusionProtocol=Supervised, Specific Modules=Segformer, YOLOS, Examples Num=4, #Params=48.12024.03 | — | — | — | — | 48.1 | |
| Uni3D3D Representation=Point Cloud2026.01 | — | — | — | — | 61.72 | |
| UniGS3D Representation=3DGS2026.01 | — | — | — | — | 68.92 | |
| X-Decoder-TProtocol=Zero-Shot Transfer, Specific Modules=FocalNet, Encoder, Examples Num=4, #Params=165M2024.03 | — | — | — | — | 34.5 |