Object Detection on VisDrone 2019 (val)
55.5AP@0.5TinyFormer-X-PBM
Evaluation Results
| Method | Links | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| TinyFormer-X-PBMParams (M)=51.5, FLOPs (G)=163.9, Input Size=640 x 6402026.05 | 55.5 | — | — | — | 34.7 | — | — | — | — | — | — | — | — | — | — | — | |
| DEIMv2-XParams (M)=50.2, FLOPs (G)=151.3, Input Size=640 x 6402026.05 | 52.3 | — | — | — | 32.2 | — | — | — | — | — | — | — | — | — | — | — | |
| TinyFormer-L-PBMParams (M)=33.6, FLOPs (G)=105.6, Input Size=640 x 6402026.05 | 52.1 | — | — | — | 32.2 | — | — | — | — | — | — | — | — | — | — | — | |
| SO-DETR-R50Params (M)=44.4, FLOPs (G)=161.4, Input Size=640 x 6402026.05 | 51.5 | — | — | — | 31.5 | — | — | — | — | — | — | — | — | — | — | — | |
| D-FINE-XParams (M)=61.5, FLOPs (G)=202.2, Input Size=640 x 6402026.05 | 51.5 | — | — | — | 31.6 | — | — | — | — | — | — | — | — | — | — | — | |
| D-FINE-LParams (M)=30.6, FLOPs (G)=90.7, Input Size=640 x 6402026.05 | 50.1 | — | — | — | 30.3 | — | — | — | — | — | — | — | — | — | — | — | |
| TinyFormer-M-PBMParams (M)=20.1, FLOPs (G)=63.0, Input Size=640 x 6402026.05 | 50 | — | — | — | 30.9 | — | — | — | — | — | — | — | — | — | — | — | |
| DEIMv2-LParams (M)=32.1, FLOPs (G)=96.4, Input Size=640 x 6402026.05 | 49.9 | — | — | — | 30.4 | — | — | — | — | — | — | — | — | — | — | — | |
| RT-DETRv2-XParams (M)=76.0, FLOPs (G)=259.0, Input Size=640 x 6402026.05 | 49.1 | — | — | — | 29.5 | — | — | — | — | — | — | — | — | — | — | — | |
| SO-DETR-R18Params (M)=20.5, FLOPs (G)=64.3, Input Size=640 x 6402026.05 | 49 | — | — | — | 29.9 | — | — | — | — | — | — | — | — | — | — | — | |
| D-FINE-MParams (M)=19.1, FLOPs (G)=56.1, Input Size=640 x 6402026.05 | 48.3 | — | — | — | 29.1 | — | — | — | — | — | — | — | — | — | — | — | |
| RT-DETRv2-LParams (M)=42.7, FLOPs (G)=136.0, Input Size=640 x 6402026.05 | 48.3 | — | — | — | 29 | — | — | — | — | — | — | — | — | — | — | — | |
| DEIMv2-MParams (M)=18.1, FLOPs (G)=51.8, Input Size=640 x 6402026.05 | 48 | — | — | — | 28.9 | — | — | — | — | — | — | — | — | — | — | — | |
| TinyFormer-S-PBMParams (M)=10.8, FLOPs (G)=32.6, Input Size=640 x 6402026.05 | 47.9 | — | — | — | 28.9 | — | — | — | — | — | — | — | — | — | — | — | |
| SO-DETR-EFV2 (Distilled)Params (M)=12.1, FLOPs (G)=33.3, Input Size=640 x 6402026.05 | 47.5 | — | — | — | 28.8 | — | — | — | — | — | — | — | — | — | — | — | |
| WDFS-DETRParams/M=19.9, GFLOPs=53.7, A100 FPS=117, Nano FPS=102026.06 | 47.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AutoAssignbackbone=ResNet-502026.01 | 46 | 2.7 | 9.5 | 20.4 | 25 | 37.6 | — | — | — | — | — | — | — | — | — | — | |
| DW + RFLAbackbone=ResNet-50, RFLA=true2026.01 | 46 | 3.8 | 10.4 | 21.8 | 25.4 | 37.7 | — | — | — | — | — | — | — | — | — | — | |
| RFAssignerbackbone=ResNet-502026.01 | 45.7 | 3.2 | 9.6 | 21.4 | 26 | 39.7 | — | — | — | — | — | — | — | — | — | — | |
| YOLOv11-XParams (M)=56.8, FLOPs (G)=194.5, Input Size=640 x 6402026.05 | 45.6 | — | — | — | 28.1 | — | — | — | — | — | — | — | — | — | — | — | |
| WCDB-YOLOParams/M=19.8, GFLOPs=60.1, A100 FPS=114, Nano FPS=92026.06 | 45.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D-FINE-SParams/M=10.18, GFLOPs=24.86, A100 FPS=140, Nano FPS=202026.06 | 45.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DEIMv2-SParams (M)=9.7, FLOPs (G)=25.3, Input Size=640 x 6402026.05 | 45 | — | — | — | 26.7 | — | — | — | — | — | — | — | — | — | — | — | |
| DERNet-MParams/M=2.9, GFLOPs=29.4, A100 FPS=134, Nano FPS=162026.06 | 44.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D-FINE-SParams (M)=9.7, FLOPs (G)=24.9, Input Size=640 x 6402026.05 | 43.4 | — | — | — | 25.8 | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv11-MParams (M)=20.0, FLOPs (G)=67.7, Input Size=640 x 6402026.05 | 43.1 | — | — | — | 25.9 | — | — | — | — | — | — | — | — | — | — | — | |
| FBRT-YOLO-SParams/M=2.9, GFLOPs=22.9, A100 FPS=143, Nano FPS=222026.06 | 42.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv9-MParams (M)=20.1, FLOPs (G)=76.8, Input Size=640 x 6402026.05 | 42 | — | — | — | 25.2 | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv8-MParams (M)=25.9, FLOPs (G)=78.9, Input Size=640 x 6402026.05 | 40.7 | — | — | — | 24.6 | — | — | — | — | — | — | — | — | — | — | — | |
| Cascade R-CNN*NWD=true2021.10 | 40.3 | 2.9 | 11.1 | 22.2 | — | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv8sQMixBlocks=0, Params=11.13M, GFLOPs=28.5, Train Time=4.47h2026.04 | 40 | — | — | — | — | — | 23.6 | — | — | — | — | — | — | — | — | — | |
| QYOLOv8sQMixBlocks=2, Params=8.70M, Param Red.=21.8%, GFLOPs=24.6, Train Time=3.94h2026.04 | 39.9 | — | — | — | — | — | 23.4 | — | — | — | — | — | — | — | — | — | |
| DERNet-SParams/M=1.3, GFLOPs=13.3, A100 FPS=162, Nano FPS=222026.06 | 39.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| QYOLOv8s (v0)QMixBlocks=5, Params=6.54M, Param Red.=41.2%, GFLOPs=21.0, Train Time=3.56h2026.04 | 39.7 | — | — | — | — | — | 23.2 | — | — | — | — | — | — | — | — | — | |
| DWbackbone=ResNet-502026.01 | 39.3 | 2.9 | 8.5 | 18.9 | 23.5 | 35.6 | — | — | — | — | — | — | — | — | — | — | |
| FCOSbackbone=ResNet-502026.01 | 39.1 | 1.5 | 5.6 | 17.1 | 22.2 | 35.4 | — | — | — | — | — | — | — | — | — | — | |
| YOLOv11-SParams (M)=9.4, FLOPs (G)=21.3, Input Size=640 x 6402026.05 | 38.7 | — | — | — | 23 | — | — | — | — | — | — | — | — | — | — | — | |
| Faster R-CNN*NWD=true2021.10 | 38.5 | 3.8 | 10.2 | 21.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Cascade R-CNNNWD=false2021.10 | 38.5 | 0.5 | 6.8 | 21.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DEIM-D-Fine-SParams/M=10.18, GFLOPs=24.85952026.06 | 38.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 23.2 | |
| YOLOv9-SParams (M)=7.2, FLOPs (G)=26.7, Input Size=640 x 6402026.05 | 38.3 | — | — | — | 22.7 | — | — | — | — | — | — | — | — | — | — | — | |
| Faster R-CNNNWD=false2021.10 | 38 | 0.1 | 6.2 | 20 | — | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv8-SParams (M)=11.1, FLOPs (G)=28.5, Input Size=640 x 6402026.05 | 37.6 | — | — | — | 22.3 | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv26-MParams/M=21.8, GFLOPs=67.92026.06 | 36.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 19.15 | |
| DERNet-MParams/M=2.9, GFLOPs=29.42026.06 | 36.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 18.4 | |
| YOLOv11-MParams/M=20.0, GFLOPs=67.72026.06 | 35.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16.13 | |
| LAF-YOLOv10Resolution=640x640, Params (M)=2.3, GFLOPs=9, FPS=1312026.02 | 35.1 | — | — | — | — | — | 20.7 | 45.2 | 34.1 | — | — | — | — | — | — | — | |
| YOLOv8nQMixBlocks=0, Params=3.01M, GFLOPs=8.1, Train Time=4.44h2026.04 | 34.9 | — | — | — | — | — | 20.1 | — | — | — | — | — | — | — | — | — | |
| QYOLOv8nQMixBlocks=2, Params=2.40M, Param Red.=20.2%, GFLOPs=7.1, Train Time=3.46h2026.04 | 34.5 | — | — | — | — | — | 19.9 | — | — | — | — | — | — | — | — | — | |
| FBRT-YOLO-MParams/M=7.36, GFLOPs=58.72026.06 | 34.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 17.5 | |
| SSCWResolution=640x640, Params (M)=2.8, GFLOPs=6.9, FPS=1152026.02 | 34.3 | — | — | — | — | — | 19.4 | 44.5 | 33.2 | — | — | — | — | — | — | — | |
| PPO2026.06 | 33.9 | — | — | — | — | — | 17.7 | — | — | 17.2 | 26.7 | 54.7 | 69.4 | 42.3 | 0.02 | — | |
| TOEResolution=640x640, Params (M)=2.5, GFLOPs=6.62026.02 | 33.8 | — | — | — | — | — | 19.7 | 45 | 33.7 | — | — | — | — | — | — | — | |
| ATSS-R50-FPN-DyHeadParams/M=38.91, GFLOPs=1102026.06 | 33.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16 | |
| TOOD-R50Params/M=32.04, GFLOPs=1992026.06 | 33.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16.2 | |
| Detector Only (init.)2026.06 | 33.7 | — | — | — | — | — | 18.9 | — | — | 18.7 | 28 | 56.6 | 66.9 | 43.9 | — | — | |
| YOLOv12-MParams/M=9.23, GFLOPs=21.22026.06 | 33.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 17.4 | |
| OSDResolution=640x640, Params (M)=1.6, GFLOPs=7.9, FPS=1362026.02 | 33.4 | — | — | — | — | — | 19.1 | 43.9 | 32.5 | — | — | — | — | — | — | — | |
| D-Fine-NParams/M=3.73, GFLOPs=7.12382026.06 | 33.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16.3 | |
| MARLNetλ=0.052026.06 | 33.3 | — | — | — | — | — | 16.7 | — | — | 15.6 | 25.5 | 55.1 | 66.6 | 41.4 | 0.015 | — | |
| Faster-RCNN-R50-FPN-CIOUParams/M=41.39, GFLOPs=2082026.06 | 32.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 15.5 | |
| QYOLOv8n (v0)QMixBlocks=5, Params=1.86M, Param Red.=38.2%, GFLOPs=6.2, Train Time=5.76h2026.04 | 32.6 | — | — | — | — | — | 18.5 | — | — | — | — | — | — | — | — | — | |
| Cascade-RCNN-R50-FPNParams/M=69.29, GFLOPs=2362026.06 | 32.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 14.9 | |
| FBRT-YOLO-SParams/M=2.9, GFLOPs=22.92026.06 | 32.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16.2 | |
| YOLO11nResolution=640x640, Params (M)=2.6, GFLOPs=6.3, FPS=1632026.02 | 32.2 | — | — | — | — | — | 18.4 | 42.9 | 32.3 | — | — | — | — | — | — | — | |
| DEIM-D-Fine-NParams/M=3.73, GFLOPs=7.12382026.06 | 32.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16 | |
| DCFLbackbone=ResNet-502026.01 | 32.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOv8nResolution=640x640, Params (M)=3, GFLOPs=8.1, FPS=1282026.02 | 32.1 | — | — | — | — | — | 19 | 42.8 | 31.6 | — | — | — | — | — | — | — | |
| BGFResolution=640x640, Params (M)=2, GFLOPs=8.6, FPS=372026.02 | 32.1 | — | — | — | — | — | 19.5 | 41.5 | 32 | — | — | — | — | — | — | — | |
| GFLParams/M=32.28, GFLOPs=2062026.06 | 32.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16.4 | |
| YOLOv10nResolution=640x640, Params (M)=2.7, GFLOPs=8.2, FPS=1362026.02 | 31.8 | — | — | — | — | — | 18.5 | 42.6 | 31.7 | — | — | — | — | — | — | — | |
| DERNet-SParams/M=1.3, GFLOPs=13.32026.06 | 31.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 15.72 | |
| YOLOv12-SParams/M=9.23, GFLOPs=21.22026.06 | 31.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 15.8 | |
| YOLOv11-SParams/M=9.4, GFLOPs=21.62026.06 | 31.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 14.57 | |
| YOLOv26-SParams/M=10.0, GFLOPs=20.52026.06 | 30.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 14.57 | |
| RT-DETR-R18Params/M=20.1, GFLOPs=61.22026.06 | 30.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 14.85 | |
| RetinaNetbackbone=ResNet-502026.01 | 29.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| YOLOX-TinyParams/M=5.035, GFLOPs=7.5782026.06 | 27.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 11.6 | |
| RetinaNet-R50-FPNParams/M=36.52, GFLOPs=2102026.06 | 27.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 12 | |
| YOLOv5nResolution=640x640, Params (M)=1.9, GFLOPs=4.5, FPS=1562026.02 | 27.4 | — | — | — | — | — | 14.8 | 38.1 | 28.2 | — | — | — | — | — | — | — | |
| FBRT-YOLO-NParams/M=0.8, GFLOPs=6.72026.06 | 26.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 12.2 | |
| YOLOv12-NParams/M=2.56, GFLOPs=6.32026.06 | 25.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 13.5 | |
| YOLOv11-NParams/M=2.59, GFLOPs=6.32026.06 | 25.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 13.8 | |
| Heuristic2026.06 | 25.7 | — | — | — | — | — | 14 | — | — | 13.4 | 26.6 | 55.9 | 67.1 | 41.8 | 0.007 | — | |
| YOLOv26-NParams/M=2.5, GFLOPs=5.22026.06 | 24.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 11.8 |