Related Objects of Interest: stop, no_parking, do_not_enter, pedestrian, do_not_u_turn, parking-sign, u_turn, warning, bus, car
Top Sign Computer Vision Models
The models below have been fine-tuned for various sign detection tasks. You can try out each model in your browser, or test an edge deployment solution (i.e. to an NVIDIA Jetson). You can use the datasets associated with the models below as a starting point for building your own sign detection model.
At the bottom of this page, we have guides on how to count signs in images and videos.
by solandeer
1004 images 9 classes
by project
142 images 79 classes
Ahead only Autotoll booth Autotoll traffic lane Bicycle / tricycle route Cross-harbour taxi stand, used with Sign 82 Cycling restriction - cyclists must dismount and push their cycles End of New Territories taxis operating area End of an expressway End of bus lane End of cycling restriction End of the prohibition, restriction or warning End of ʻno stoppingʼ restriction Give way to traffic on major road Green minibus stand Keep left Keep right Lantau taxi stand, used with Sign 82 Left lane shows bus lane for franchised and other buses during the time shown Left lane shows bus lane for franchised buses only during the time shown Length over which the prohibition or hazard exists
by asdasd adsa
3252 images 21 classes
by Crickets
6558 images 52 classes
* 50% probability of horizontal flip * Auto-orientation of pixel data (with EXIF-orientation stripping) * Random rotation of between -15 and +15 degrees * Resize to 640x640 (Stretch) * annotate, and create datasets * collaborate with your team on computer vision projects * collect & organize images * export, train, and deploy computer vision models * understand and search unstructured image data * use active learning to improve your dataset over time 23 24 25 26 27 28 30 31 32 33
by college
3480 images 27 classes
* Auto-orientation of pixel data (with EXIF-orientation stripping) * Random Gaussian blur of between 0 and 2.5 pixels * Random brigthness adjustment of between -25 and +25 percent * Resize to 640x640 (Stretch) * Salt and pepper noise was applied to 14 percent of pixels * annotate, and create datasets * collaborate with your team on computer vision projects * collect & organize images * export, train, and deploy computer vision models * understand and search unstructured image data * use active learning to improve your dataset over time 22 23 24 25 26 ============================== For state of the art Computer Vision training notebooks you can use with this dataset, Road Sign Detector - v7 Add blur and noise Roboflow is an end-to-end computer vision platform that helps you
by yangbaba
3202 images 20 classes
by Objdetection
507 images 3 classes
3664 images 22 classes
by yolov4
2444 images 11 classes
by Eta
81 images 71 classes
fork Auto not permitted Big truck not permitted Caution Disabled Caution animals Caution children Caution cliff Caution falling rocks Caution road users Danger Dangerous Uphill Grade Dangerous downhill grade Keep left Keep right Lane Ends Merge left Lane Ends Merge right Low-lying road Maximum speed Motocycle not permitted Narrow bridge ahead
by Classes
3388 images 21 classes