Unofficial ONNX conversion of Roboflow RF-DETR — nano
Single-file ONNX exporting the raw model outputs (no NMS, no sigmoid, no top-k, no resizing — all pre/postprocessing stays outside the graph). Unofficial conversion; not endorsed by the upstream authors. Upstream-reported COCO AP: 48.4 (not re-measured here).
Class table is UNVERIFIED. RF-DETR emits a 91-slot DETR-family COCO layout (index 0 = background). Confirm the column count and no-object slot against this exact file (Netron + one known image) before relying on the labels.
Input contract
| name | input |
| dtype / layout | float32, NCHW |
| shape | [1, 3, 384, 384] (static) |
Preprocessing: RGB → resize (stretch) to 384×384 → float32 ÷255 → ImageNet mean/std after ÷255 → HWC→CHW → batch dim.
Output contract
| output | shape | meaning |
|---|---|---|
labels |
[1, 300, 91] |
raw class logits (apply sigmoid yourself) |
dets |
[1, 300, 4] |
normalized cxcywh |
91-class COCO (torchvision 91-slot), 300 queries, DETR-style set prediction — no NMS.
Artifact
model.onnx SHA-256 |
d5082b3a93eba673cf71763842be73fb852ab79962dfb303bbf0012ab8805059 (114,731,473 bytes) |
License
Apache-2.0 (see LICENSE / NOTICE). Upstream: Roboflow RF-DETR.
- Downloads last month
- 9