Parameters1.5M
GFLOPs5.2
Input Size640px
Best mAP38.5%
LicenseApache-2.0
Architecture
Type
transformer
Backbone
DINOv3-distilled ViT
Neck
HybridEncoder
Head
DEIM
Benchmark Results
Performance on COCO val2017 across different hardware configurations
| Hardware | Runtime | mAP@50-95 | FPS | Latency | VRAM |
|---|---|---|---|---|---|
| NVIDIA RTX 5070 Ti | PyTorch FP32 | 38.5% | 51.6 | 19.4ms | 61 MB |
Speed Breakdown(NVIDIA RTX 5070 Ti)
5.5ms
13.1ms
0.8ms
Preprocess
Inference
Postprocess (NMS)
Usage with LibreYOLO
from libreyolo import LIBREYOLO
# Load model (auto-downloads from HuggingFace if not found locally)
model = LIBREYOLO("libredeimv2pico.pth")
# 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: 38.5% mAP