No abstract

Keyword Spotting (KWS) enables speech-based user interaction on smart devices. Always-on and battery-powered application scenarios for smart devices put constraints on hardware resources and power consumption, while also demanding high accuracy as well as real-time capability. Previous architectures first extracted acoustic features and then applied a neural network to classify keyword probabilities, optimizing towards memory footprint and execution time.Compared to previous publications, we took additional steps to reduce power and memory consumption without reducing classification accuracy. Power-consuming audio preprocessing and data transfer steps are eliminated by directly classifying from raw audio. For this, our end-to-end architecture extracts spectral features using parametrized Sinc-convolutions. Its memory footprint is further reduced by grouping depthwise separable convolutions. Our network achieves the competitive accuracy of 96.4% on Google's Speech Commands test set with only 62k parameters.

No abstract

No abstract

Efficient error correction and key derivation is a prerequisite to generate secure and reliable keys from PUFs. The most common methods can be divided into linear schemes and pointer-based schemes. This work compares the performance of several previous designs on an algorithmic level concerning the required number of PUF response bits, helper data bits, number of clock cycles, and FPGA slices for two scenarios. One targets the widely used key error probability of 10 −6 , while the other one requires a key error probability of 10 −9. In addition, we provide a wide span of new implementation results on state-of-the-art Xilinx FPGAs and set them in context to old synthesis results on legacy FPGAs.