<p>This study presents a novel biometric approach to identify operators, given only streams of their control movements within a manual control task setting. In the present task subjects control a simulated, remotely operated robotic arm, attempting to dock onto a satellite in orbit. The proposed methodology utilizes the Hierarchical Temporal Memory (HTM) algorithm to distinguish operators by their unique control behaviors. Results presented compare the identification performance of HTM with Dynamic Time Warping (DTW) and Edit Distance on Real Sequences (EDR), in both static and real-time data settings. The HTM method outperformed both DTW and EDR in the real- time setting, and matched DTW in the static setting. Observed superior performance of the HTM algorithm lays the foundation for the extension of the proposed methodology to other motion- monitoring applications, such as real-time workload assessment, motion/simulator sickness onset or distraction detection.</p><p><br></p><p>The data gathered in the study was posted to IEEE-dataport, DOI: <a href="http://dx.doi.org/10.21227/wpyf-r927" target="_blank">10.21227/wpyf-r927</a><br></p><div><br></div>
Temporal Memory Network Towards Real-Time Video Understanding