We explore the use of machine learning to detect under-rotation in figure skating jumps. Under- rotation in jumps is difficult for the skater to sense but learning to recognize under-rotation is an impor- tant part of learning proper jump technique. To address this difficulty, we present the Under-rotation Monitor, or UR Monitor, a system for detecting under-rotated figure skating jumps in real-time. UR Monitor uses a single inertial measurement unit (IMU) attached to the skater's waist that sends a stream of accelerometer and gyroscope data to a mobile phone via Bluetooth. The mobile phone creates and sends an input vector of each jump to a web-hosted API that returns a response from our trained classifier indicating whether it considered that jump as 'under-rotated', or 'completed rotation'. The classifier is trained and tested on a collection of 444 jumps, of which only 121 are under-rotated. We also present a process for addressing an imbalanced dataset on which the classifier trains. Our classifier achieves an F1-score of only 0.66, suggesting that noise and imbalance in the data set are significant issues.

The task of the judge of difficulty in trampoline gymnastics is to check the elements and difficulty values entered on the competition cards and the difficulty of each element according to a numeric system. To do this, the judge must count all somersaults and twists for each jump during a routine and thus record the difficulty of the routine. This assessment can be automated with the help of inertial measurement units (IMUs) and facilitate the judges’ task during the competition. Currently, there is no known reliable method for the automated detection and recognition of the various elements to determine the difficulty of an exercise in trampoline gymnastics. Accordingly, a total of 2076 jumps and 50 different jump types were recorded over the course of several training sessions. In the first instance, 10 different jump types were used to train different machine learning (ML) models. Eight ML models were used for the automatic jump classification. Supervised learning approaches include a naive classifier, deep feedforward neural network, convolutional neural network, k‑nearest neighbors, Gaussian naive Bayes, support-vector classification, gradient boosting classifier, and stochastic gradient descent. When all classifiers were compared for accuracy, i.e., how many jumps were correctly detected by the ML model, the deep feedforward neural network and the convolutional neural network provided the best matches with 96.4 and 96.1%, respectively. The findings of this study will help to develop the automated classification of sensor-based data to support the judge and, simultaneously, for automated training logging.