Jérémie Decouchant
42 records found
1
This thesis paper addresses the vulnerability of Deep Neural Networks (DNNs) to adversarial attacks. We introduce Multi-Scale Inpainting Defense (MSID), a novel adversarial purification method leveraging a pre-trained diffusion denoising probabilistic model (DDPM) for targeted pe
...
Testing Byzantine Fault Tolerant Algorithms
Evaluating the correctness of Tendermint protocol using ByzzFuzz
The reliability of Byzantine Fault Tolerant (BFT) consensus protocols is critical for the robustness of modern distributed systems, i.e., in blockchain technologies. Testing of BFT protocols is crucial, as consequences of faults in their implementation can lead to malicious users
...
Although Byzantine Fault Tolerant (BFT) protocols such as HotStuff are nominally resistant to a number of faulty or unreliable participants, implementation or design errors can cause violations in their expected properties. Because of this, it is useful to have reliable automated
...
Byzantine fault-tolerant protocols have been around for decades, offering the guarantee of agreement on a correct value even in the presence of arbitrary failures. These protocols have become a critical part of achieving consensus in distributed systems and are widely used nowada
...
Byzantine Fault Tolerant (BFT) protocols are designed to achieve consensus even in the presence of Byzantine faults. Although BFT protocols provide strong theoretical guarantees, bugs in the implementation of the protocols can allow for malicious activity. While previous work, li
...
Fuzzing has been a popular approach in the domain of software testing due to its efficiency and capability to uncover unexpected bugs. Fuzz testing was originally developed in the days of sequential programs. With the rise of multi-core devices and increasing demand for computat
...
Conflict-free replicated data types (CRDTs) offer high-availability low-latency updates to data without the need for central coordination. Despite the current vast collection of CRDTs, few works have been done on maintaining probabilistic membership information using CRDTs. In th
...
The Vesper Protocol
Leveraging Zero-Knowledge Proofs and SGX Enclaves in Hyperledger Fabric Smart Contracts
This work explores the feasibility of combining zero-knowledge proofs with SGX enclave protection technology, using the Hyperledger fabric, as the testing environment. The focus is on assessing the viability of this combination in real-world scenarios where post-quantum security
...
Federated learning (FL) allows multiple clients to train a machine learning model on a server without sharing their private data. To reach a consensus, the server collects alternative information such as model updates. The sub-field of heterogeneous FL investigates scenarios wher
...
In the modern digital landscape, cybersecurity threats are a significant concern, particularly for publicly accessible computer systems. Vulnerabilities, or flaws in system design, can be exploited by malicious actors to compromise system security and integrity. This paper explor
...
Go With The Flow: Fault-Tolerant Decentralized Training of Large Language Models
Decentralised Training of Large Language Models
Motivated by the emergence of Large Language Models (LLMs) and the importance of democratizing their training, we propose Go With The Flow, the first practical decentralized training framework for LLMs. Differently from existing distributed and federated training frameworks, Go W
...
Federated Learning (FL) is a decentralized machine learning approach that provides a privacy-friendly way of training models by keeping the datasets of participating parties private. Some challenges FL faces are the lack of incentives to encourage participation in the learning pr
...
Software development often relies on dependencies managed by package managers to simplify the integration of external libraries and frameworks, reducing development time. However, developers sometimes choose to bundle dependencies directly within their software packages. Bundling
...
Optimizing Database Joins
Cost Models and Benchmarking for CPU and GPU Systems
Optimizing SQL query execution through effective cost models is a critical challenge in database management systems (DBMS). This thesis introduces a modular benchmarking system for cost models, with a pluggable architecture for both cost models and execution engines, enabling com
...
Small embedded devices are becoming more prevalent in the world with each passing year to improve our quality of life. However, as more devices are created, an increasing number of older devices are declared obsolete despite still being used. This results in an increasing amount
...
Leveraging Feature Extraction to Detect Adversarial Examples
Let's Meet in the Middle
Previous research has explored the detection of adversarial examples with dimensional reduction and Out-of-Distribution (OOD) recognition. However, these approaches are not effective against white-box adversarial attacks. Moreover, recent OOD methods that utilize hidden units hin
...
In recent years, the rapid advancements in big data, machine learning, and artificial intelligence have led to a corresponding rise in privacy concerns. One of the solutions to address these concerns is federated learning. In this thesis, we will look at the setting of vertical f
...
In federated learning systems, a server maintains a global model trained by a set of clients based on their local datasets. Conventional synchronous FL systems are very sensitive to system heterogeneity since the server needs to wait for the slowest clients in each round. Asynchr
...
Combining SAT solvers with heuristic ideas for solving RCPSP with logical constraints
An exploration of variable ordering heuristics impact on solving RCPSP-log
This paper provides a novel method of solving the resource-constrained project scheduling problem (RCPSP) with logical constraints (RCPSP-log) using satisfiability (SAT) solving and integrating variable selection heuristics. The extension provides two additional precedences: OR c
...
Why Midas would be a terrible secretary
Using a greedy approach to enhance SAT for the Preemptive Resource-Constrained project scheduling problem with set up time
This paper presents a new greedy heuristic to extend SAT Solvers when solving the Preemptive resource-constrained project scheduling problem (PRCPSP-ST). The heuristic uses domain-specific knowledge to generate a fixed order of variable selection. We also extend previous work int
...