Object Detection on GEN1 (mAP 0.5:0.95)
0.39mAP (0.5:0.95)SpikeYOLO
Evaluation results
| Method | Links | |
|---|---|---|
| SpikeYOLOPlatform=GPU, Activation Unit=LIF, Resolution=640 × 640, # Activation Units (M)=84, # Parameters (M)=41, Inference Rate (Sa/s)=312026.04 | 0.39 | |
| EMS-YOLO Res34Regime=gradient-trained, GPU, Params.=14.4 M, Latency=7 ms2026.05 | 0.31 | |
| YOLOv8nPlatform=Orin Nano, Activation Unit=SiLU, Resolution=300 × 300, # Activation Units (M)=16.81, # Parameters (M)=3.16, Inference Rate (Sa/s)=84, Dynamic Energy (mJ)=15.542026.04 | 0.29 | |
| EMS-YOLO Res18Regime=gradient-trained, GPU, Params.=9.34 M, Latency=7 ms2026.05 | 0.286 | |
| Model-2Platform=Loihi 2, Activation Unit=LIF, Resolution=256 × 192, # Activation Units (M)=0.768, # Parameters (M)=3.83, Inference Rate (Sa/s)=97, Dynamic Energy (mJ)=6.402026.04 | 0.23 | |
| SSD-Densenet121Platform=GPU, Activation Unit=pLIF, Resolution=300 × 300, # Activation Units (M)=52, # Parameters (M)=8.2, Inference Rate (Sa/s)=392026.04 | 0.19 |