A5023968611- Cryptography and Data Security
- Privacy-Preserving Technologies in Data
- Cryptographic Implementations and Security
- Coding theory and cryptography
- Chaos-based Image/Signal Encryption
- Complexity and Algorithms in Graphs
- Interconnection Networks and Systems
- Parallel Computing and Optimization Techniques
- Cancer Genomics and Diagnostics
- Embedded Systems Design Techniques
- Cryptography and Residue Arithmetic
- Internet Traffic Analysis and Secure E-voting
- Distributed and Parallel Computing Systems
- VLSI and Analog Circuit Testing
- VLSI and FPGA Design Techniques
- Advanced Data Storage Technologies
- Genomic variations and chromosomal abnormalities
- Cloud Data Security Solutions
- History and Theory of Mathematics
- Analytic Number Theory Research
- Physical Unclonable Functions (PUFs) and Hardware Security
- Blockchain Technology Applications and Security
- Advanced Steganography and Watermarking Techniques
- Scheduling and Optimization Algorithms
- Security and Verification in Computing
University of Lausanne
2019-2023
CEA LIST
2010-2021
Commissariat à l'Énergie Atomique et aux Énergies Alternatives
2011-2021
École Polytechnique Fédérale de Lausanne
2020-2021
Integra (United States)
2011-2015
Université de Technologie de Compiègne
2010-2011
Centre National de la Recherche Scientifique
2010-2011
Heuristics and Diagnostics for Complex Systems
2011
In this work we present Armadillo a compilation chain used for compiling applications written in high-level language (C++) to on encrypted data. The back-end of the is based homomorphic encryption. tool-chain further automatically handle huge amount parallelism so as mitigate performance overhead using
Genotype imputation is a fundamental step in genomic data analysis, where missing variant genotypes are predicted using the existing of nearby "tag" variants. Although researchers can outsource genotype imputation, privacy concerns may prohibit genetic sharing with an untrusted service. Here, we developed secure efficient homomorphic encryption (HE) techniques. In HE-based methods, while it transit, at rest, and analysis. It only be decrypted by owner. We compared three state-of-the-art...
Abstract Background Privacy-preserving computations on genomic data, and more generally medical is a critical path technology for innovative, life-saving research to positively equally impact the global population. It enables algorithms be securely deployed in cloud because operations encrypted databases are conducted without revealing any individual genomes. Methods secure computation have shown significant performance improvements over last several years. However, it still challenging...
The use of remote services offered by cloud providers have been popular in the last lustrum. Services allow users to store files, or analyze data for several purposes, like health-care message analysis. However, when personal are sent Cloud, may lose privacy on data-content, and other side those their own businesses. In this paper, we present our solution health-data directly into Cloud while preserving privacy. Our makes homomorphic encryption protect during particular, developed a mobile...
The ever-growing number of cores in embedded chips emphasizes more than ever the complexity inherent to parallel pro- gramming. To solve these programmability issues, there is a renewed interest dataflow paradigm. In this context, we present compilation toolchain for ΣC language, which allows hierarchical construction stream applications and automatic mapping application an manycore target. As demonstration toolchain, implementation H.264 encoder evaluate its performance on Kalray's MPPA chip.
We present a framework GenoPPML for privacy-preserving machine learning in the context of sensitive genomic data processing. The technology combines secure multiparty computation techniques based on recently proposed Manticore model training and fully homomorphic encryption TFHE inference. was successfully used to solve breast cancer prediction problems gene expression datasets coming from distinct private sources while preserving their privacy - solution winning 1st place both Tracks I III...
ABSTRACT Genotype imputation is a fundamental step in genomic data analysis such as GWAS, where missing variant genotypes are predicted using the existing of nearby ‘tag’ variants. Imputation greatly decreases genotyping cost and provides high-quality estimates common genotypes. As population panels increase, e.g., TOPMED Project, genotype becoming more accurate, but it requires high computational power. Although researchers can outsource imputation, privacy concerns may prohibit genetic...
One of the 3 tracks iDASH Privacy & Security Workshop 2017 competition was to execute a whole genome variants search on private genomic data. Particularly, application find top most significant SNPs (Single-Nucleotide Polymorphisms) in database records labeled with control or case. In this paper we discuss solution submitted by our team competition.Privacy and confidentiality data had be ensured using Intel SGX enclaves. The typical use-case is multi-party computation (each party possessing...
In 2009, Craig Gentry introduced the first "fully" homomorphic encryption scheme allowing arbitrary circuits to be evaluated on encrypted data. Homomorphic is a very powerful cryptographic primitive, though it has often been viewed by practitioners as too inefficient for practical applications. However, performance of these schemes come long way from that Gentry's original work: there are now several well-maintained libraries implementing and protocols demonstrating impressive results,...
In this work we describe a message packing and unpacking method for homomorphic ciphertexts. Messages are packed into the coefficients of plaintext polynomials. We propose an procedure which allows to obtain ciphertext each message. The ciphertexts represents solution reducing transmission bottleneck in cloud based applications, particular when sending calculations results. results (packing ratio, time) compared existing methods on trans-ciphering.
This paper deals with semantics-preserving parallelism reduction methods for cyclo-static dataflow applications.Parallelism is the process of equivalent actors fusioning.The principal objectives are to decrease memory footprint an application and increase its execution performance.We focus on methodologies constrained by throughput.A generic methodology introduced.Experimental results provided asserting performance proposed method.
In this paper, we examine and propose a solution for the challenges of sharing genome sequence data querying on cloud server in per-sonalized medicine scenarios. We develop privacy-preserving, secure efficient personalized medicine. The that making use stream cipher-based homomorphic transciphering server, to show effectiveness scenario. This paper also provides comparative analysis well-known existing encryption solutions BGV FV schemes combined with FLIP cipher demonstrate efficiency...
FPGA floorplannig consists in finding a satisfactory placement of the different pre-determined regions design, onto resource matrix that composes hardware. Performance is ensured by minimizing distances between communicating regions, as well and their I/O ports on FPGA. To this very challenging problem, additional features can be added, such taking into account existence partially-reconfigurable regions. This paper presents solution method we proposed RAW Floorplanning Desing Contest,...
Abstract Classification algorithms/tools become more and powerful pervasive. Yet, for some use cases, it is necessary to be able protect data privacy while benefiting from the functionalities they provide. Among tools that may used ensure such privacy, we are focusing in this paper on functional encryption . These relatively new cryptographic primitives enable evaluation of functions over encrypted inputs, outputting cleartext results. Theoretically, property makes them well-suited process...