English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Ball trajectory data are one of the most fundamental and useful informationin the evaluation of players' performance and analysis of game strategies.Although vision-based object tracking techniques have been developed to analyzesport competition videos, it is still challenging to recognize and position ahigh-speed and tiny ball accurately. In this paper, we develop a deep learningnetwork, called TrackNet, to track the tennis ball from broadcast videos inwhich the ball images are small, blurry, and sometimes with afterimage tracksor even invisible. The proposed heatmap-based deep learning network is trainedto not only recognize the ball image from a single frame but also learn flyingpatterns from consecutive frames. TrackNet takes images with a size of$640\times360$ to generate a detection heatmap from either a single frame orseveral consecutive frames to position the ball and can achieve high precisioneven on public domain videos. The network is evaluated on the video of themen's singles final at the 2017 Summer Universiade, which is available onYouTube. The precision, recall, and F1-measure of TrackNet reach $99.7\%$,$97.3\%$, and $98.5\%$, respectively. To prevent overfitting, 9 additionalvideos are partially labeled together with a subset from the previous datasetto implement 10-fold cross-validation, and the precision, recall, andF1-measure are $95.3\%$, $75.7\%$, and $84.3\%$, respectively. A conventionalimage processing algorithm is also implemented to compare with TrackNet. Ourexperiments indicate that TrackNet outperforms conventional method by a bigmargin and achieves exceptional ball tracking performance. The dataset and demovideo are available at https://nol.cs.nctu.edu.tw/ndo3je6av9/.

TrackNet: A Deep Learning Network for Tracking High-speed and Tiny Objects in Sports Applications