Skip to content

Commit 19ab4aa

Browse files
SkalskiPSkalskiP
authored
Merge pull request roboflow#308 from roboflow/docs/better-benchmark-formatting-and-parameter-count
Added parameter counts and standardized architecture naming in the benchmark table
2 parents e38c093 + 4a37374 commit 19ab4aa

File tree

2 files changed

+32
-32
lines changed

2 files changed

+32
-32
lines changed

‎README.md‎

Lines changed: 16 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -34,22 +34,22 @@ The table below shows the performance of RF-DETR medium, compared to comparable
3434

3535
![rf-detr-coco-rf100-vl-9](https://media.roboflow.com/rfdetr/pareto1.png)
3636

37-
|family|size |coco_map50|coco_map50@95|rf100vl_map50|rv100vl_map50@95|latency|
38-
|------|------|----------|------------|-------------|---------------|-------|
39-
|RF-DETR|Nano |67.6 |48.4 |84.1 |57.1 |2.32 |
40-
|RF-DETR|Small |72.1 |53.0 |85.9 |59.6 |3.52 |
41-
|RF-DETR|Medium|73.6 |54.7 |86.6 |60.6 |4.52 |
42-
|YOLO11|n |52.0 |37.4 |81.4 |55.3 |2.49 |
43-
|YOLO11|s |59.7 |44.4 |82.3 |56.2 |3.16 |
44-
|YOLO11|m |64.1 |48.6 |82.5 |56.5 |5.13 |
45-
|YOLO11|l |65.3 |50.2 |x |x |6.65 |
46-
|YOLO11|x |66.5 |51.2 |x |x |11.92 |
47-
|LW-DETR|Tiny |60.7 |42.9 |x |x |1.91 |
48-
|LW-DETR|Small |66.8 |48.0 |84.5 |58.0 |2.62 |
49-
|LW-DETR|Medium|72.0 |52.6 |85.2 |59.4 |4.49 |
50-
|D-FINE |Nano |60.2 |42.7 |83.6 |57.7 |2.12 |
51-
|D-FINE |Small |67.6 |50.7 |84.5 |59.9 |3.55 |
52-
|D-FINE |Medium|72.6 |55.1 |84.6 |60.2 |5.68 |
37+
| Architecture | COCO AP<sub>50</sub> | COCO AP<sub>50:95</sub> | RF100VL AP<sub>50</sub> | RF100VL AP<sub>50:95</sub> | Latency (ms) | Params (M) |
38+
|:------------:|:--------------------:|:--------------------------:|:--------------------------:|:---------------------------:|:---------------:|-------------:|
39+
| RF-DETR-N | 67.6 | 48.4 | 84.1 | 57.1 | 2.32 | 30.5 |
40+
| RF-DETR-S | 72.1 | 53.0 | 85.9 | 59.6 | 3.52 | 32.1 |
41+
| RF-DETR-M | 73.6 | 54.7 | 86.6 | 60.6 | 4.52 | 33.7 |
42+
| YOLO11-N | 52.0 | 37.4 | 81.4 | 55.3 | 2.49 | 2.6 |
43+
| YOLO11-S | 59.7 | 44.4 | 82.3 | 56.2 | 3.16 | 9.4 |
44+
| YOLO11-M | 64.1 | 48.6 | 82.5 | 56.5 | 5.13 | 20.1 |
45+
| YOLO11-L | 65.3 | 50.2 | x | x | 6.65 | 25.3 |
46+
| YOLO11-X | 66.5 | 51.2 | x | x | 11.92 | 56.9 |
47+
| LW-DETR-T | 60.7 | 42.9 | x | x | 1.91 | 12.1 |
48+
| LW-DETR-S | 66.8 | 48.0 | 84.5 | 58.0 | 2.62 | 14.6 |
49+
| LW-DETR-M | 72.0 | 52.6 | 85.2 | 59.4 | 4.49 | 28.2 |
50+
| D-FINE-N | 60.2 | 42.7 | 83.6 | 57.7 | 2.12 | 3.8 |
51+
| D-FINE-S | 67.6 | 50.7 | 84.5 | 59.9 | 3.55 | 10.2 |
52+
| D-FINE-M | 72.6 | 55.1 | 84.6 | 60.2 | 5.68 | 19.2 |
5353

5454
[See our benchmark notes in the RF-DETR documentation.](https://rfdetr.roboflow.com/learn/benchmarks/)
5555

‎docs/learn/benchmarks.md‎

Lines changed: 16 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -6,21 +6,21 @@ The table below shows the performance of RF-DETR, compared to other object detec
66

77
![rf-detr-coco-rf100-vl-9](https://media.roboflow.com/rfdetr/pareto1.png)
88

9-
|family|size |coco_map50|coco_map5095|rf100vl_map50|rv100vl_map5095|latency|
10-
|------|------|----------|------------|-------------|---------------|-------|
11-
|RF-DETR|Nano |67.6 |48.4 |84.1 |57.1 |2.32 |
12-
|RF-DETR|Small |72.1 |53.0 |85.9 |59.6 |3.52 |
13-
|RF-DETR|Medium|73.6 |54.7 |86.6 |60.6 |4.52 |
14-
|YOLO11|n |52.0 |37.4 |81.4 |55.3 |2.49 |
15-
|YOLO11|s |59.7 |44.4 |82.3 |56.2 |3.16 |
16-
|YOLO11|m |64.1 |48.6 |82.5 |56.5 |5.13 |
17-
|YOLO11|l |65.3 |50.2 |x |x |6.65 |
18-
|YOLO11|x |66.5 |51.2 |x |x |11.92 |
19-
|LW-DETR|Tiny |60.7 |42.9 |x |x |1.91 |
20-
|LW-DETR|Small |66.8 |48.0 |84.5 |58.0 |2.62 |
21-
|LW-DETR|Medium|72.0 |52.6 |85.2 |59.4 |4.49 |
22-
|D-FINE |Nano |60.2 |42.7 |83.6 |57.7 |2.12 |
23-
|D-FINE |Small |67.6 |50.7 |84.5 |59.9 |3.55 |
24-
|D-FINE |Medium|72.6 |55.1 |84.6 |60.2 |5.68 |
9+
| Architecture | COCO AP<sub>50</sub> | COCO AP<sub>50:95</sub> | RF100VL AP<sub>50</sub> | RF100VL AP<sub>50:95</sub> | Latency (ms) | Params (M) |
10+
|:------------:|:--------------------:|:--------------------------:|:--------------------------:|:---------------------------:|:---------------:|-------------:|
11+
| RF-DETR-N | 67.6 | 48.4 | 84.1 | 57.1 | 2.32 | 30.5 |
12+
| RF-DETR-S | 72.1 | 53.0 | 85.9 | 59.6 | 3.52 | 32.1 |
13+
| RF-DETR-M | 73.6 | 54.7 | 86.6 | 60.6 | 4.52 | 33.7 |
14+
| YOLO11-N | 52.0 | 37.4 | 81.4 | 55.3 | 2.49 | 2.6 |
15+
| YOLO11-S | 59.7 | 44.4 | 82.3 | 56.2 | 3.16 | 9.4 |
16+
| YOLO11-M | 64.1 | 48.6 | 82.5 | 56.5 | 5.13 | 20.1 |
17+
| YOLO11-L | 65.3 | 50.2 | x | x | 6.65 | 25.3 |
18+
| YOLO11-X | 66.5 | 51.2 | x | x | 11.92 | 56.9 |
19+
| LW-DETR-T | 60.7 | 42.9 | x | x | 1.91 | 12.1 |
20+
| LW-DETR-S | 66.8 | 48.0 | 84.5 | 58.0 | 2.62 | 14.6 |
21+
| LW-DETR-M | 72.0 | 52.6 | 85.2 | 59.4 | 4.49 | 28.2 |
22+
| D-FINE-N | 60.2 | 42.7 | 83.6 | 57.7 | 2.12 | 3.8 |
23+
| D-FINE-S | 67.6 | 50.7 | 84.5 | 59.9 | 3.55 | 10.2 |
24+
| D-FINE-M | 72.6 | 55.1 | 84.6 | 60.2 | 5.68 | 19.2 |
2525

2626
_We are actively working on RF-DETR Large and X-Large models using the same techniques we used to achieve the strong accuracy that RF-DETR Medium attains. This is why RF-DETR Large and X-Large is not yet reported on our pareto charts and why we haven't benchmarked other models at similar sizes. Check back in the next few weeks for the launch of new RF-DETR Large and X-Large models._

0 commit comments

Comments
 (0)