• An energy-efficient and error-resilient server ecosystem exceeding conservative scaling limits 

      Karakonstantis G., Tovletoglou K., Mukhanov L., Vandierendonck H., Nikolopoulos D.S., Lawthers P., Koutsovasilis P., Maroudas M., Antonopoulos C.D., Kalogirou C., Bellas N., Lalis S., Venugopal S., Prat-Pérez A., Lampropulos A., Kleanthous M., Diavastos A., Hadjilambrou Z., Nikolaou P., Sazeides Y., Trancoso P., Papadimitriou G., Kaliorakis M., Chatzidimitriou A., Gizopoulos D., Das S. (2018)
      The explosive growth of Internet-connected devices will soon result in a flood of generated data, which will increase the demand for network bandwidth as well as compute power to process the generated data. Consequently, ...
    • Optimum use of existing sensor information for gas turbine diagnostics 

      Stamatis, A. G. (2008)
      Gas Path Analysis, (GPA), has been proven a powerful tool for Gas Turbine fault detection and isolation. Restriction on GPA effectiveness is mainly due to limited instrumentation imposed by sensor technology and plant ...
    • Resource-adaptive deep learning for visual speech recognition 

      Koumparoulis A., Potamianos G., Thomas S., da Silva Morais E. (2020)
      We focus on the problem of efficient architectures for lipreading that allow trading-off computational resources for visual speech recognition accuracy. In particular, we make two contributions: First, we introduce ...