Semantic Segmentation on DSEC (test)
74.94mIoUBRENet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BRENetModality=RGB-Event, Backbone=MiT-B2, Event Representation=MET2025.05 | 74.94 | 95.85 | |
| HFR (Ideal)Input=I_{t+δt}, Causal (CS)=false, Anytime (AT)=false2026.03 | 73.91 | — | |
| LiFR-SegInput=I_t, E_{t-Δt→t+δt}, Causal (CS)=true, Anytime (AT)=true2026.03 | 73.82 | — | |
| EISNetModality=RGB-Event, Backbone=MiT-B2, Event Representation=AET2025.05 | 73.07 | 95.12 | |
| SegFormerModality=RGB, Backbone=MiT-B22025.05 | 71.99 | 94.97 | |
| SegNeXtModality=RGB, Backbone=MSCAN-B2025.05 | 71.55 | 94.89 | |
| CMNeXtInput=I_t, E_{t→t+δt}, Causal (CS)=true, Anytime (AT)=true2026.03 | 70.13 | — | |
| SE-AdapterModality=RGB-Event, Backbone=SAM, Event Representation=MSP2025.05 | 69.77 | 93.58 | |
| TLX + Seg.Input=I_t, I_{t+Δt}, E..., Causal (CS)=false, Anytime (AT)=true2026.03 | 68.17 | — | |
| EISNetInput=I_t, E_{t→t+δt}, Causal (CS)=true, Anytime (AT)=true2026.03 | 68.11 | — | |
| LFR (Baseline)Input=I_t, Causal (CS)=true, Anytime (AT)=false2026.03 | 67.67 | — | |
| GEPBackbone=ViT-B/14, Pre-training Dataset=Event-1.8M, Ep.=242026.03 | 67.37 | 76.27 | |
| CMNeXtModality=RGB-Event, Backbone=MiT-B2, Event Representation=Voxel Grid2025.05 | 67.2 | 93.13 | |
| Hybrid-SegModality=RGB-Event, Backbone=FCN, Event Representation=Voxel Grid2025.05 | 66.57 | 94.27 | |
| CMXModality=RGB-Event, Backbone=MiT-B2, Event Representation=Voxel Grid2025.05 | 65.29 | 92.61 | |
| GEPBackbone=ViT-S/14, Pre-training Dataset=Event-1.8M, Ep.=242026.03 | 65.22 | 74.66 | |
| TESPECBackbone=Swin-T, Pre-training Dataset=1Mpx [42]2026.03 | 62.77 | 70.61 | |
| GEPBackbone=ViT-S/14, Pre-training Dataset=N-ImageNet, Ep.=242026.03 | 60.43 | 68.76 | |
| CMESSModality=RGB-Event, Backbone=E2ViD, Event Representation=Voxel Grid2025.05 | 59.53 | 91.11 | |
| SpikingEDNModality=RGB-Event, Backbone=FCN, Event Representation=Voxel Grid2025.05 | 58.32 | — | |
| OpenESSModality=RGB-Event, Backbone=E2VID, Event Representation=Voxel Grid2025.05 | 57.21 | 90.21 | |
| EDCNet-S2DModality=RGB-Event, Backbone=ResNet-101, Event Representation=Voxel Grid2025.05 | 56.75 | 92.39 | |
| ECDDPBackbone=ViT-S/16, Pre-training Dataset=E-TartanAir, Ep.=3002026.03 | 56.38 | 66 | |
| Spike-BRGNetModality=RGB-Event, Backbone=MiT-B2, Event Representation=Voxel Grid2025.05 | 54.95 | — | |
| ESSDataset=Cityscape, Ep.=502026.03 | 53.3 | 62.94 | |
| ECDPBackbone=ViT-S/16, Pre-training Dataset=N-ImageNet, Ep.=3002026.03 | 52.52 | 60.55 | |
| HALSIEModality=RGB-Event, Backbone=FCN, Event Representation=Voxel Grid2025.05 | 52.43 | 89.01 | |
| DINOv2Backbone=ViT-S/14, Pre-training Dataset=LVD-142M2026.03 | 52.17 | 59.8 | |
| MAEBackbone=ViT-B/16, Pre-training Dataset=ImageNet1K, Ep.=8002026.03 | 51.96 | 59.84 | |
| BeiTBackbone=ViT-B/16, Pre-training Dataset=ImageNet1K, Ep.=8002026.03 | 51.9 | 59.66 | |
| EV-SegNetModality=Event, Backbone=Xception, Event Representation=6-Channel Image2025.05 | 51.76 | 88.61 | |
| ESSModality=Event, Backbone=E2ViD, Event Representation=Voxel Grid2025.05 | 51.57 | 89.25 | |
| MEMBackbone=ViT-S/16, Pre-training Dataset=N-ImageNet, Ep.=752026.03 | 44.62 | 51.39 | |
| ViTBackbone=ViT-S/162026.03 | 32.66 | 40.84 |