Publikationen

2022

Joint Classification and Trajectory Regression of Online Handwriting using a Multi-Task Learning Approach

Felix Ott, David Rügamer, Lucas Heublein, Bernd Bischl, Christopher Mutschler

In: WACV 2022

Multivariate Time Series (MTS) classification is important in various applications such as signature verification, person identification, and motion recognition. In deep learning these classification tasks are usually learned using the cross-entropy loss. A related yet different task is predicting trajectories observed as MTS. Important use cases include handwriting reconstruction, shape analysis, and human pose estimation. The goal is to align an arbitrary dimensional time series with its ground truth as accurately as possible while reducing the error in the prediction with a distance loss and the variance with a similarity loss. Although learning both losses with Multi-Task Learning (MTL) helps to improve trajectory alignment, learning often remains difficult as both tasks are contradictory. We propose a novel neural network architecture for MTL that notably improves the MTS classification and trajectory regression performance in online handwriting (OnHW) recognition. We achieve this by jointly learning the cross-entropy loss in combination with distance and similarity losses. On an OnHW task of handwritten characters with multivariate inertial and visual data inputs we are able to achieve crucial improvements (lower error with less variance) of trajectory prediction while still improving the character classification accuracy in comparison to models trained on the individual tasks.

Searching for Soccer Scenes using Siamese Neural Networks

Luca Reeb

In: Towards Data Science 2022

We have access to a large soccer database, containing a seasons worth of tracking-data, i.e. player trajectories, game statistics and expert-annotated events like pass or shot at goal from the German Bundesliga. While events allow you to find set-pieces like corner-kicks, the results are coarsely grained in that they do not consider how the players acted during the event. Also, some situations of potential interest, like counter attack, are not represented by an event. To enable fine-grained analysis of soccer matches, player movement (i.e. tracking-data) has to be considered.