v1.0
版本发布时间: 2020-06-26 03:25:46
ultralytics/yolov5最新发布版本:v7.0(2022-11-22 23:29:09)
YOLOv5 1.0 Release Notes
- June 22, 2020: PANet updates: increased layers, reduced parameters, faster inference and improved mAP 364fcfd.
- June 19, 2020: FP16 as new default for smaller checkpoints and faster inference d4c6674.
- June 9, 2020: CSP updates: improved speed, size, and accuracy. Credit to @WongKinYiu for excellent CSP work.
- May 27, 2020: Public release of repo. YOLOv5 models are SOTA among all known YOLO implementations.
- April 1, 2020: Start development of future YOLOv3/YOLOv4-based PyTorch models in a range of compound-scaled sizes.
Pretrained Checkpoints
Model | APval | APtest | AP50 | SpeedGPU | FPSGPU | params | FLOPS | |
---|---|---|---|---|---|---|---|---|
YOLOv5s | 36.6 | 36.6 | 55.8 | 2.1ms | 476 | 7.5M | 13.2B | |
YOLOv5m | 43.4 | 43.4 | 62.4 | 3.0ms | 333 | 21.8M | 39.4B | |
YOLOv5l | 46.6 | 46.7 | 65.4 | 3.9ms | 256 | 47.8M | 88.1B | |
YOLOv5x | 48.4 | 48.4 | 66.9 | 6.1ms | 164 | 89.0M | 166.4B | |
YOLOv3-SPP | 45.6 | 45.5 | 65.2 | 4.5ms | 222 | 63.0M | 118.0B |
** APtest denotes COCO test-dev2017 server results, all other AP results in the table denote val2017 accuracy.
** All AP numbers are for single-model single-scale without ensemble or test-time augmentation. Reproduce by python test.py --img 736 --conf 0.001
** SpeedGPU measures end-to-end time per image averaged over 5000 COCO val2017 images using a GCP n1-standard-16 instance with one V100 GPU, and includes image preprocessing, PyTorch FP16 image inference at --batch-size 32 --img-size 640, postprocessing and NMS. Average NMS time included in this chart is 1-2ms/img. Reproduce by python test.py --img 640 --conf 0.1
** All checkpoints are trained to 300 epochs with default settings and hyperparameters (no autoaugmentation).
1、 Arial.Bold.ttf 733.38KB
2、 Arial.Narrow.Bold.ttf 180.1KB
3、 Arial.Narrow.ttf 175.29KB
4、 Arial.Rounded.Bold.ttf 48.14KB
5、 Arial.ttf 755.11KB
6、 Arial.Unicode.ttf 22.2MB
7、 banner-api.png 1.14MB
8、 banner-export-competition.png 198.29KB
9、 banner-social.jpg 233.77KB
10、 banner-survey.png 144.92KB
11、 banner_hub_1920.png 189.49KB
12、 banner_hub_3840.png 593.75KB
13、 bus.jpg 476.01KB
14、 caltech101.zip 130.85MB
15、 caltech256.zip 1.06GB
16、 cifar10.zip 139.89MB
17、 cifar100.zip 173.96MB
18、 cifar100_nested.zip 175.85MB
19、 coco128-seg.zip 6.79MB
20、 coco128-segments.zip 6.79MB
21、 coco128.zip 6.66MB
22、 coco128_with_yaml.zip 6.68MB
23、 coco2017labels-segments-sama.zip 199.4MB
24、 coco2017labels-segments.zip 168.51MB
25、 coco2017labels.zip 46.39MB
26、 coco2017val.zip 780.41MB
27、 decelera_landscape.mov 5.12MB
28、 decelera_portrait.mov 3.16MB
29、 fashion-mnist.zip 47.01MB
30、 GlobalWheat2020_labels.zip 2.77MB
31、 image-contributors-1280.png 741.41KB
32、 image-contributors-1920.png 1.4MB
33、 image-contributors.svg 2.11MB
34、 image-integrations-loop.png 255.48KB
35、 image-local_logging.jpg 470.17KB
36、 imagenette.zip 1.45GB
37、 imagenette160.zip 102.56MB
38、 imagenette320.zip 331.63MB
39、 imagewoof.zip 1.25GB
40、 imagewoof160.zip 96.19MB
41、 imagewoof320.zip 319MB
42、 logo-aws-small.png 6.4KB
43、 logo-clearml.png 14.88KB
44、 logo-colab-small.png 11.96KB
45、 logo-comet.png 17.76KB
46、 logo-deci.png 3.81KB
47、 logo-docker-small.png 3.66KB
48、 logo-gcp-small.png 7.3KB
49、 logo-gradient.png 76.59KB
50、 logo-kaggle-small.png 2.47KB
51、 logo-neuralmagic.png 13.4KB
52、 logo-paperspace.png 20.58KB
53、 logo-roboflow-long.png 20.68KB
54、 logo-roboflow.png 6.99KB
55、 logo-social-facebook.png 655B
56、 logo-social-github.png 2.23KB
57、 logo-social-instagram.png 2.05KB
58、 logo-social-linkedin.png 1.18KB
59、 logo-social-producthunt.png 1.89KB
60、 logo-social-twitter.png 1.76KB
61、 logo-social-youtube.png 1.1KB
62、 logo-transparent.png 163B
63、 logo-wb-long.png 9.87KB
64、 logo-wb-small.png 4.25KB
65、 logo-wb.png 5.77KB
66、 mnist.zip 28.24MB
67、 mnist160.zip 70.04KB
68、 mnist2560.zip 1.05MB
69、 model_comparison.png 118.14KB
70、 model_plot.png 233.01KB
71、 results.png 264.85KB
72、 splash-export-competition.png 289.85KB
73、 splash.jpg 789.58KB
74、 splash_horizontal.png 320.11KB
75、 splash_vertical.png 823.98KB
76、 VisDrone2019-DET-test-challenge.zip 292.02MB
77、 VisDrone2019-DET-test-dev.zip 296.83MB
78、 VisDrone2019-DET-train.zip 1.44GB
79、 VisDrone2019-DET-val.zip 77.86MB
80、 VOCtest_06-Nov-2007.zip 418.01MB
81、 VOCtrainval_06-Nov-2007.zip 425.26MB
82、 VOCtrainval_11-May-2012.zip 1.82GB
83、 yolov5l.pt 91.57MB
84、 yolov5m.pt 41.87MB
85、 yolov5s.pt 14.46MB
86、 yolov5x.pt 170.15MB
87、 zidane.jpg 164.99KB