Identification for Human-Machine Interaction

Automation can improve the performance and maintain the safety of a human operated system. Unsurprisingly, dependendence upon automation has increased dramatically; however, the existing state of automation has restricted applicability since it is unable to operate safely in arbitrary circumstances.
Our group is investigating how to expand and improve human-machine collaboration by developing semi-autonomous architectures. In these systems, the burden of control is shared between a human who is predominantly in charge of controlling the machine and an autonomous controller who intervenes in the operation of the human when they are distracted and the controller is capable and required to preserve safety. Our group has developed tools from algebraic topology that are able to perform robust pose tracking from multiple cameras to construct these semi-autonomous architectures. By taking advantage of this pose tracking method and other types of sensors and processing our tools, our group is currently developing stochastic models of human operation of machines. 

References

  1. V. Shia, Y. Gao, R. Vasudevan, K. Campbell, T. Lin, F. Borrelli and R. Bajcsy, “Semiautonomous Vehicular Control Using Driver Modeling,” IEEE Transactions on Intelligent Transportation Systems, vol. 15, no. 6, pp. 2696-2709, November 2014. [url]

  2. K. Campbell, V. Shia, R. Vasudevan and R. Bajcsy, “Probabilistic Driver Models for Semiautonomous Vehicles,” in Digital Signal Processing for In-Vehicle Systems, 2013.

  3. R. Vasudevan, V. Shia, Y. Gao, R. Cervera-Navarro, R. Bajcsy and F. Borrelli, “Safe semi-autonomous control with enhanced driver modeling,” in American Control Conference, pp. 2896-2903, 2012. [pdf]

  4. E. Lobaton, R. Vasudevan, R. Alterovitz and R. Bajcsy, “Robust topological features for deformation invariant image matching,” in International Conference on Computer Vision, pp. 2516-2523, 2011. [pdf]
  5. E. Lobaton, R. Vasudevan, R. Bajcsy and R. Alterovitz, “Local occlusion detection under deformations using topological invariants,” in European Conference on Computer Vision, pp. 101-114, 2010. [pdf]