Existing multi-camera solutions for automatic scorekeeping in steel-tip darts are very expensive and thus inaccessible to most players . We present a new approach to keypointdetection and apply it to predict dart scores from a single image taken fromany camera angle . Because DeepDarts relies only on single images, it has the potential to be deployed onedge devices, giving anyone with a smartphone access to an automatic dartscoring system for steel tip darts . The code and datasets are available for download and the code and dataset is available . In a second more challenging dataset containing limited training data(830 images) and various camera angles, we utilize transfer learning andextensive data augmentation to achieve a test accuracy of 84.0% . In the primary datasetcontaining 15k images captured from a face-on view of the dartboard using asmartphone, DeepDars predicted the total score correctly in 94.7% of the . training data (830 images), we utilize a secondMore challenging dataset with limited training and various training data, we use a second More challenging dataset (830 pictures) and . various camera . data augmented to achieve an accuracy of . 84. 0% . The data augmentmentmentments are available .
Author(s) : William McNally, Pascale Walters, Kanav Vats, Alexander Wong, John McPheeLinks : PDF - Abstract
Code :
Keywords : data - camera - training - dataset - images -
- Khan Academy Multivariable Calculus
- Learning How to Learn on Coursera
- fast.ai Machine Learning
- Mathematics for Machine Learning by Marc Peter Deisenroth, A. Aldo Faisal and Cheng Soon Ong
- Google Cloud Professional Data Engineer Specialization
- Deeplearning.ai Specialization by Andrew Ng
- Data Scientist with Python
- Khan Academy Linear Algebra