Ludwig Kürzinger

Ludwig Kürzinger

Auto-generated profile, based on publicly available information

Ludwig Kürzinger

Chair of Human-Machine Communication, Technical University of Munich, Munich, Germany

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 September 2020
Verified
Book Chapter
Closed access
LKLudwig Kürzinger
Ludwig Kürzinger
DWDominik Winkelbauer
Dominik Winkelbauer
LLLujun Li
Lujun Li
TWTobias Watzel
Tobias Watzel
GRGerhard Rigoll
Gerhard Rigoll

11 May 2020 in Journal of Cryptographic Engineering
Verified
Journal Article
Open access
PDF file
MHMatthias Hiller
Matthias Hiller
LKLudwig Kürzinger
Ludwig Kürzinger
GSGeorg Sigl
Georg Sigl
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 $$10^{-6}$$ 10 - 6 , while the other one requires a key error probability of $$10^{-9}$$ 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.

1 May 2020
Verified
Conference Paper
Closed access
SMSimon Mittermaier
Simon Mittermaier
LKLudwig Kurzinger
Ludwig Kurzinger
BWBernd Waschneck
Bernd Waschneck
GRGerhard Rigoll
Gerhard Rigoll
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.

20 January 2016
Verified
Conference Paper
Open access
PDF file
VIVincent Immler
Vincent Immler
MHMaxim Hennig
Maxim Hennig
LKLudwig Kürzinger
Ludwig Kürzinger
GSGeorg Sigl
Georg Sigl
This work deals with key generation based on Physically Obfuscated Keys (POKs), i.e., a certain type of tamperevident Physical Unclonable Function (PUF) that can be used as protection against invasive physical attacks. To design a protected device, one must take attacks such as probing of data lines or penetration of the physical security boundary into consideration. For the implementation of a POK as a countermeasure, physical properties of a material - which covers all parts to be protected - are measured. After measuring these properties, i.e. analog values, they have to be quantized in order to derive a cryptographic key. This paper will present and discuss the impact of the quantization method with regard to three parameters: key quality, tamper-sensitivity, and reliability. Our contribution is the analysis of two different quantization schemes considering these parameters. Foremost, we propose a new approach to achieve improved tampersensitivity in the worst-case with no information leakage

30 June 2021 in Electronics
Verified
Journal Article
Open access
PDF file
LLLujun Li
Lujun Li
Zhenxing LuZhenxing Lu
Zhenxing Lu
TWTobias Watzel
Tobias Watzel
LKLudwig Kürzinger
Ludwig Kürzinger
GRGerhard Rigoll
Gerhard Rigoll
Generative adversarial networks (GANs) have shown their superiority for speech enhancement. Nevertheless, most previous attempts had convolutional layers as the backbone, which may obscure long-range dependencies across an input sequence due to the convolution operator’s local receptive field. One popular solution is substituting recurrent neural networks (RNNs) for convolutional neural networks, but RNNs are computationally inefficient, caused by the unparallelization of their temporal iterations. To circumvent this limitation, we propose an end-to-end system for speech enhancement by applying the self-attention mechanism to GANs. We aim to achieve a system that is flexible in modeling both long-range and local interactions and can be computationally efficient at the same time. Our work is implemented in three phases: firstly, we apply the stand-alone self-attention layer in speech enhancement GANs. Secondly, we employ locality modeling on the stand-alone self-attention layer. Lastly, we investigate the functionality of the self-attention augmented convolutional speech enhancement GANs. Systematic experiment results indicate that equipped with the stand-alone self-attention layer, the system outperforms baseline systems across classic evaluation criteria with up to 95% fewer parameters. Moreover, locality modeling can be a parameter-free approach for further performance improvement, and self-attention augmentation also overtakes all baseline systems with acceptably increased parameters.

1 July 2015
Verified
Conference Paper
Closed access
MHMatthias Hiller
Matthias Hiller
LKLudwig Kurzinger
Ludwig Kurzinger
GSGeorg Sigl
Georg Sigl
SMSven Muelich
Sven Muelich
SPSven Puchinger
Sven Puchinger
MBMartin Bossert
Martin Bossert
Physical Unclonable Functions (PUFs) enable secure key storage for integrated circuits and FPGAs. PUF responses are noisy such that error correction is required to generate stable cryptographic keys. One popular approach is to use error-correcting codes. We present an area-optimized VLSI implementation of a recent Generalized Concatenated (GC) code construction using Reed-Muller codes. Reed-Muller codes have the advantage that there exist very efficient decoders. Our new Reed decoding implementation makes extensive use of a circular shift register. The functionality is extended so that it can also handle erasure symbols to improve the error correction capability. The overall GC code decoder occupies less than 110 slices and two block RAMs on an entry-level FPGA, and has a key error probability of 1.5 × 10 -9 . The slice count is reduced by 50% compared to the reference implementation.