Research

This research paper presents BioFast, a privacy-preserving face verification protocol (a system for confirming someone's identity using their face while keeping their facial data secret) that uses FHE (fully homomorphic encryption, a type of encryption that lets computers process data without decrypting it first) and cryptographic hash functions (mathematical functions that scramble data in a one-way process). The protocol improves on existing methods by introducing more efficient packing techniques for processing multiple calculations at once and replacing interactive garbled circuits (cryptographic methods requiring back-and-forth communication) with a non-interactive comparison scheme, resulting in faster and more efficient face verification.

Team workspaces like Google Workspace and Microsoft OneDrive allow third-party applications called add-ons to extend their functionality, but researchers found that these add-ons can bypass permission isolation (access control rules set by administrators) in multi-user environments, potentially letting them escalate privileges (gain higher access levels than intended). A testing tool called TAI discovered 41 problematic interactions that could allow permission escalation across these platforms.

Distributed learning (training AI models across multiple computers while keeping data in different locations) faces three types of security threats: attackers reconstructing data from gradients (mathematical updates), inferring information from the final model, and manipulating the training process. Fortress is a framework that combines three defensive techniques—secure aggregation (encrypted combination of data), differential privacy (adding noise to protect individual data), and malicious resilience (verification to catch cheating)—while reducing communication overhead by 6.2× to 32.3× compared to existing approaches.

This paper presents LMIRA, a blockchain-based framework designed to improve data sharing between vehicles across different network domains in the Internet of Vehicles (IoV, a system where vehicles communicate with each other and infrastructure). The framework addresses three main problems: uneven distribution of computational load on the blockchain, unreliable shared vehicle information, and slow authentication processes that threaten privacy. LMIRA uses edge computing (processing data closer to where it's generated rather than in distant data centers) with a prediction model to anticipate transaction volume, evaluates vehicle data quality using multiple factors, and introduces a new secure authentication method.

This paper addresses how to make UAV (unmanned aerial vehicle) communications more energy-efficient while keeping the transmitted information hidden from malicious detectors. The researchers developed a method to jointly optimize the UAV's transmit power and flying location to maximize covert energy efficiency (the amount of information successfully delivered per unit of energy used) while maintaining secrecy against an adversary with imperfect detection capabilities.

Verifiable Encrypted Timed Signatures (VETS) is a new cryptographic tool that improves security in blockchain transactions by separating the ability to verify a signature from the ability to extract it. Unlike earlier timed signature schemes where someone could start solving immediately upon receiving the signature, VETS ensures extraction only happens after an unlock key is released, which can be tied to an external condition like publishing a refund transaction, protecting users in atomic swaps (trades between two parties without a middleman) from malicious actors with powerful hardware.

SLIMQ is a privacy-preserving system for searching multi-dimensional datasets (data with multiple attributes) stored on cloud servers without revealing the search query or sensitive data to the cloud. The system uses R-trees (a data structure for organizing spatial data), prefix encoding (converting queries into searchable patterns), and secret sharing (splitting information among multiple parties) to perform efficient searches while keeping data private.

This research proposes a method for performing (α, β)-core queries, which identify tightly connected groups of nodes in bipartite graphs (networks with two types of connected entities), while keeping the underlying data encrypted. The approach uses structured encryption (a technique that allows computations on encoded data without decoding it) and avoids processing the entire graph by using index tables and graph coloring to efficiently find relevant nodes, while supporting queries on both node and edge attributes with acceptable performance overhead.

Graph unlearning removes specific information from graph neural networks (GNNs, which are AI models that process data organized as networks of connected nodes). In machine learning-as-a-service (MLaaS, where companies host AI models for users to access), service providers usually cannot see the original training data, making existing unlearning methods impractical. This paper introduces SCGU (subgraph-based certified graph unlearning), a method that lets service providers directly modify model parameters to remove specific information without needing access to the training data, using only a smaller portion of the model related to what needs to be removed.

Compromised credential checking (C3) services like HaveIBeenPwned let users check if their passwords were exposed in data breaches, but existing services don't verify that the server is telling the truth, meaning a malicious C3 server could falsely claim a breached password is safe. This paper proposes VerC3, a new system that adds verification to C3 services by using two servers instead of one, signing breached password records with the data owner's digital signature, and defining protocols so clients can confirm the server responded honestly.

Federated learning (a machine learning approach where multiple computers train a model together without sharing raw data) systems using TEEs (trusted execution environments, secure areas within processors that protect data even from the main operating system) can be attacked by malicious servers that exploit TEE limitations like state rollback (reverting to previous states) and I/O manipulation (tampering with input/output). This paper presents DisT-FL, a solution using multiple TEEs working together in an append-only ledger (an unchangeable record of transactions) to secure federated learning aggregation (combining results from all participants) and prevent these attacks.

Bhra-VITARIT is a new atomic swap protocol (a method for exchanging digital assets between parties without needing to trust each other) that works on blockchains with limited scripting capabilities, like Bitcoin. Unlike existing protocols, it supports weighted participation (where different nodes have different levels of authority) and uses threshold signatures (requiring approval from a minimum number of servers out of a larger group) to enable secure exchanges while keeping on-chain computations minimal.

This research paper presents a system for generating synthetic data (artificially created images and text) to train Text-Based Person Retrieval models, which match people in images to written descriptions of them. The authors created a pipeline that generates diverse synthetic person images and automatically writes descriptions for them, without needing real photos, and tested whether models trained only on this synthetic data work as well as those trained on real images in different real-world situations.

Object detectors used in safety-critical systems are vulnerable to physical adversarial patch attacks (images or objects designed to fool AI vision systems), which work by disrupting how information flows through layers of the neural network rather than simply corrupting visual features. The paper proposes CaDe, a defense method that uses hierarchical stability constraints and adaptive sparse causal decoupling to stop adversarial perturbations from propagating through the network layers, improving detection accuracy by 5-6% against these attacks while maintaining real-time performance.

Adversarial training (a method where AI models learn to defend against adversarial attacks, which are inputs designed to fool the model) typically requires expensive multi-step calculations, so researchers use single-step versions to save computing power. However, single-step methods create some misleading adversarial examples, called "fakers," that actually weaken the model instead of strengthening it. This paper proposes FAST (Faker-Alleviating Single-step adversarial Training), which reduces the impact of fakers during training by dynamically adjusting how the model learns from difficult examples and introducing auxiliary samples to stabilize the training process.

Identity-based matchmaking encryption (IB-ME) is a privacy technique that lets two parties securely communicate with mutual access control, but it has a major problem: a central authority can create encryption and decryption keys for any user, enabling key abuse and making it hard to hold anyone accountable. This paper proposes accountable-authority IB-ME (A-IB-ME), which adds accountability by allowing the creator of encryption or decryption keys to be identified if disputes arise, while maintaining the privacy and security features of regular IB-ME.

Federated unlearning (FUL, a process that removes a user's data influence from machine learning models trained across multiple computers) helps protect privacy by letting users exercise their right to be forgotten. This paper proposes a new framework where an FL server uses game theory (a mathematical approach to modeling strategic decision-making) and prospect theory (a model of how people make decisions under uncertainty) to incentivize clients to keep more data during unlearning while preventing selfish behavior.

This research addresses security vulnerabilities in mmWave (millimeter wave, a type of wireless communication using very high frequencies) MIMO systems during beam alignment, where directional beams used to establish connections are susceptible to impersonation attacks (where an attacker tricks a device by pretending to be someone else). The researchers propose two authentication schemes that use physical layer features like antenna coupling patterns and signal arrival angles to verify device identity and prevent these attacks.

Machine unlearning is a process that allows AI models to forget the influence of specific training samples, which is important for privacy and safety. Researchers tested whether existing unlearning methods actually work when the training dataset contains similar samples to the ones being removed, and found that most methods fail to completely eliminate a target sample's influence even when compared to retraining from scratch (rebuilding the model from the beginning with the unwanted sample excluded).