Parameters33.7M
GFLOPs0.0
Input Size576px
Best mAP57.4%
LicenseApache-2.0
Architecture
Type
transformer
Backbone
DINOv2
Neck
Agg2Former
Head
DETR
Benchmark Results
Performance on COCO val2017 across different hardware configurations
| Hardware | Runtime | mAP@50-95 | FPS | Latency | VRAM |
|---|---|---|---|---|---|
| NVIDIA Jetson Orin Nano Super 8GB | ONNX Runtime FP32 | 57.4% | 4.9 | 204.7ms | - |
| NVIDIA Jetson Orin Nano Super 8GB | PyTorch FP32 | 57.4% | 5.7 | 174.5ms | 193 MB |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP16 | 57.4% | 12.5 | 79.8ms | - |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP32 | 57.4% | 11.6 | 85.9ms | - |
| NVIDIA RTX 5070 Ti | ONNX Runtime FP32 | 57.4% | 59.8 | 16.7ms | - |
| NVIDIA RTX 5070 Ti | PyTorch FP32 | 57.3% | 30.3 | 33.0ms | 194 MB |
| NVIDIA RTX 5070 Ti | TensorRT FP16 | 57.4% | 62.9 | 15.9ms | - |
| NVIDIA RTX 5070 Ti | TensorRT FP32 | 57.4% | 63.7 | 15.7ms | - |
Speed Breakdown(NVIDIA Jetson Orin Nano Super 8GB)
15.6ms
153.4ms
5.5ms
Preprocess
Inference
Postprocess (NMS)
Usage with LibreYOLO
from libreyolo import LibreYOLO
# Load model (auto-downloads from HuggingFace if not found locally)
model = LibreYOLO("LibreRFDETRm.pt")
# Run inference
result = model("image.jpg", conf=0.25, iou=0.45)
# Process results
print(f"Found {len(result)} objects")
print(result.boxes.xyxy) # bounding boxes (N, 4)
print(result.boxes.conf) # confidence scores (N,)
print(result.boxes.cls) # class IDs (N,)detrnms-freePaper: 54.7% mAP