Predicting Shot Success for Table Tennis Using Video Analysis and Machine Learning

Draschkowitz, Lukas; Draschkowitz, Christoph; Hlavacs, Helmut

doi:10.1007/978-3-319-08189-2_2

Cited by 2 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We observe the poorest performance from [41] as it doesn't posses any capacity to oversee player or scene specific context. The model neither incorporates historical player behaviour nor the ball trajectory information when predicting the shot outcome.…”

Section: Validationmentioning

confidence: 90%

“…The model neither incorporates historical player behaviour nor the ball trajectory information when predicting the shot outcome. The baseline LSTM model incorporates this information, and gains a significant performance boost compared to [41]. We would like to compare it against the model of Wei et.…”

Section: Validationmentioning

confidence: 99%

“…This model utilises speed, ball bounce location, player feet locations and a set of hand crafted dominance features to classify the return shot type. In the next baseline we adapt the classifier model proposed in [41] for classifying the shot success in table tennis matches. Essentially the model utilises hits (number of shots that the player of interest has played thus far from the beginning of the rally), a shot quality variable computed by speed and bounce position of the incoming shot, shot direction, player ids, and current player points.…”

Section: Baselinesmentioning

confidence: 99%

See 2 more Smart Citations

Memory Augmented Deep Generative models for Forecasting the Next Shot Location in Tennis

Fernando

Denman

Sridharan

et al. 2020

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

This paper presents a novel framework for predicting shot location and type in tennis. Inspired by recent neuroscience discoveries we incorporate neural memory modules to model the episodic and semantic memory components of a tennis player. We propose a Semi Supervised Generative Adversarial Network architecture that couples these memory models with the automatic feature learning power of deep neural networks, and demonstrate methodologies for learning player level behavioural patterns with the proposed framework. We evaluate the effectiveness of the proposed model on tennis tracking data from the 2012 Australian Tennis open and exhibit applications of the proposed method in discovering how players adapt their style depending on the match context.

show abstract

Section: Validationmentioning

confidence: 90%

Section: Validationmentioning

confidence: 99%

Section: Baselinesmentioning

confidence: 99%

See 1 more Smart Citation