Semantic Segmentation on DDD17
78.56mIoUBRENet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BRENetParams (M)=37.69, MACS (G)=16.372025.05 | 78.56 | — | |
| MambaSegPublication=Ours, Modality=Image-Event, Backbone=Mamba, Representation=Voxel Grid2025.12 | 77.56 | 96.33 | |
| MambaSegBackbone=Mamba, Params (M)=25.44, MACs (G)=15.592025.12 | 77.56 | — | |
| EISNetParams (M)=34.39, MACS (G)=17.32025.05 | 75.03 | — | |
| EISNetPublication=TMM'24, Modality=Image-Event, Backbone=Transformer, Representation=AET2025.12 | 75.03 | 96.04 | |
| EISNetBackbone=Transformer, Params (M)=34.39, MACs (G)=17.302025.12 | 75.03 | — | |
| SwiftFormerBackbone=Transformer, Params (M)=64.61, MACs (G)=23.882025.12 | 74.31 | — | |
| SegFormerBackbone=Transformer, Params (M)=51.54, MACs (G)=19.782025.12 | 73.25 | — | |
| CMNeXtPublication=CVPR'23, Modality=Image-Event, Backbone=Transformer, Representation=Voxel Grid2025.12 | 72.67 | 95.74 | |
| CMNeXtBackbone=Transformer, Params (M)=58.68, MACs (G)=16.322025.12 | 72.67 | — | |
| CMXPublication=TITS'23, Modality=Image-Event, Backbone=Transformer, Representation=Voxel Grid2025.12 | 71.88 | 95.64 | |
| CMXBackbone=Transformer, Params (M)=66.56, MACs (G)=16.292025.12 | 71.88 | — | |
| SegNeXtPublication=NeurIPS'21, Modality=Image, Backbone=CNN, Representation=-2025.12 | 71.46 | 95.97 | |
| SegFormerPublication=NeurIPS'22, Modality=Image, Backbone=Transformer, Representation=-2025.12 | 71.05 | 95.73 | |
| SE-AdapterPublication=ICRA'24, Modality=Image-Event, Backbone=Transformer, Representation=MSP2025.12 | 69.06 | 95.32 | |
| CMXParams (M)=66.56, MACS (G)=16.292025.05 | 67.47 | — | |
| Hybrid-SegPublication=AAAI'25, Modality=Image-Event, Backbone=CNN+SNN, Representation=Voxel Grid2025.12 | 67.31 | 95.07 | |
| CMNeXtParams (M)=58.68, MACS (G)=16.322025.05 | 66.99 | — | |
| ScaleEventBackbone=ViT-L/16, Evaluation Protocol=Fully-Supervised2026.03 | 65.08 | 92.62 | |
| STPBackbone=Swin-T/7, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=ICCV’252026.03 | 63.29 | — | |
| OpenESSBackbone=E2VID, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=CVPR’242026.03 | 63 | 91.05 | |
| ScaleEventBackbone=ViT-B/16, Evaluation Protocol=Fully-Supervised2026.03 | 62.81 | 92.21 | |
| ECDDPBackbone=Swin-T/7, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=ECCV’242026.03 | 62.56 | — | |
| KWYAFBackbone=MiT-B0, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=AAAI’252026.03 | 62.41 | 91.32 | |
| STPBackbone=ResNet-50, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=ICCV’252026.03 | 62.13 | — | |
| FC-CLIPBackbone=CNeXt-L, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=NeurIPS’232026.03 | 62.01 | 90.68 | |
| EDCNetPublication=TITS'22, Modality=Image-Event, Backbone=CNN, Representation=Voxel Grid2025.12 | 61.99 | 93.8 | |
| EDCNetBackbone=CNN, Params (M)=23.06, MACs (G)=6.142025.12 | 61.99 | — | |
| ESSPublication=ECCV'22, Modality=Event, Backbone=CNN, Representation=Voxel Grid2025.12 | 61.37 | 91.08 | |
| ESS-SupBackbone=E2VID, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=ECCV’222026.03 | 61.37 | 91.08 | |
| MaskCLIPBackbone=ViT-B/16, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=ECCV’222026.03 | 61.27 | 90.5 | |
| HALSIEPublication=WACV'24, Modality=Image-Event, Backbone=CNN+SNN, Representation=Voxel Grid2025.12 | 60.66 | 92.5 | |
| HALSIEBackbone=SNN-ANN, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=WACV’242026.03 | 60.66 | 92.5 | |
| ESSPublication=ECCV'22, Modality=Image-Event, Backbone=CNN, Representation=Voxel Grid2025.12 | 60.43 | 90.37 | |
| ESEGBackbone=MiT-B1, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=AAAI’252026.03 | 59.97 | 90.68 | |
| ScaleEventBackbone=ViT-S/16, Evaluation Protocol=Fully-Supervised2026.03 | 59.64 | 91.39 | |
| ECDPBackbone=ResNet-50, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=ICCV’232026.03 | 59.15 | — | |
| DTLBackbone=ResNet-50, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=ICCV’212026.03 | 58.8 | — | |
| EVDistillBackbone=ResNet-34, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=CVPR’212026.03 | 58.02 | — | |
| OpenESSBackbone=ResNet-50, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=CVPR’242026.03 | 57.01 | — | |
| EDCNet-S2DParams (M)=23.06, MACS (G)=6.142025.05 | 56.75 | — | |
| ECDDPBackbone=ViT-S/16, Pre-training Strategy=Event Pretraining, Evaluation Protocol=Fully-Supervised, Present at=ECCV’242026.03 | 55.73 | — | |
| EV-SegNetPublication=CVPR'19, Modality=Event, Backbone=CNN, Representation=6-Channel2025.12 | 54.81 | 89.76 | |
| Ev-SegNetBackbone=CNN, Params (M)=29.09, MACs (G)=73.622025.12 | 54.81 | — | |
| Ev-SegNetBackbone=Xception, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=CVPRW’192026.03 | 54.81 | 89.76 | |
| EvSegformerBackbone=MiT-B1, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=TIP’232026.03 | 54.41 | 94.72 | |
| PVT-FPNBackbone=ResNet-34, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=ICCV’212026.03 | 53.89 | 94.28 | |
| ESSBackbone=E2VID, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=ECCV’222026.03 | 53.09 | 88.43 | |
| EV-SegNetParams (M)=29.092025.05 | 51.76 | — | |
| E2VIDBackbone=ResNet-18, Pre-training Strategy=RGB initialization, Evaluation Protocol=Fully-Supervised, Present at=TPAMI’192026.03 | 48.47 | 85.84 |