Open Access

The majority of anterior cruciate ligament (ACL) injuries in team sports are non-contact injuries, with cutting maneuvers identified as high-risk tasks. Young female handball players have been shown to be at greater risk for ACL injuries than males. One risk factor for ACL injuries is the magnitude of the knee abduction moment (KAM). Cutting technique variables on foot placement, overall approach and knee kinematics have been shown to influence the KAM. Since injury risk is believed to increase with increasing task complexity, the purpose of the study was to test the effect of task complexity on technique variables that influence the KAM in female handball players during fake-and-cut tasks.

Every new technology is used by us humans almost without hesitation. Usually the military use comes first. Examples from recent history are the use of chemical weapons by Germany in the First World War and of atomic bombs in the Second World War by the US. Now, with the rapid advances in microelectronics over the past few decades, a wave of its application, called digitization, is spreading around the world with barely any control mechanisms. In many areas this has simplified and enriched our lives, but it has also encouraged abuse. The adaptation of legislation to contain the obvious excesses of “digitization” such as hate mail and anonymous threats is lagging behind massively. We hear almost nothing about technology assessment through systematic research; it is demanded at most by a few, usually small groups in civil society, which draw attention to the threats to humankind—future and present—and the Earth's ecosystem. One such group, the Federation of German Scientists (VDW) e.V., in the spirit of the responsibility of science for the peaceful and considered application of the possibilities it creates, asked three of its study groups to jointly organize its 2019 Annual Conference. The study groups “Health in Social Change,” “Education and Digitization,” and “Technology Assessment of Digitization” formulated the following position paper for the 2019 VDW Annual Conference, entitled “Ambivalences of the Digital.”

Convolutional neural networks (CNN) define the state-of-the-art solution on many perceptual tasks. However, current CNN approaches largely remain vulnerable against adversarial perturbations of the input that have been crafted specifically to fool the system while being quasi-imperceptible to the human eye. In recent years, various approaches have been proposed to defend CNNs against such attacks, for example by model hardening or by adding explicit defence mechanisms. Thereby, a small “detector” is included in the network and trained on the binary classification task of distinguishing genuine data from data containing adversarial perturbations. In this work, we propose a simple and light-weight detector, which leverages recent findings on the relation between networks’ local intrinsic dimensionality (LID) and adversarial attacks. Based on a re-interpretation of the LID measure and several simple adaptations, we surpass the state-of-the-art on adversarial detection by a significant margin and reach almost perfect results in terms of F1-score for several networks and datasets. Sources available at: https://github.com/adverML/multiLID