Existing blockchain systems scale poorly because of their distributed consensus protocols. Current attempts at improving scalability are limited to cryptocurrency. Scaling under general workloads (i.e., non-cryptocurrency applications) remains an open question. This work takes a principled approach apply sharding in order improve transaction throughput scale. is challenging, however, due the fundamental difference failure models between databases and blockchain. To achieve our goal, we first...

The wide application of deep learning technique has raised new security concerns about the training data and test data. In this work, we investigate model inversion problem under adversarial settings, where adversary aims at inferring information target model's from prediction values. We develop a solution to train second neural network that acts as inverse perform inversion. can be trained with black-box accesses model. propose two main techniques towards in settings. First, leverage...

This paper presents a new direct Fourier-based algorithm for performing image-to-image registration to subpixel accuracy, where the image differences are restricted translations and uniform changes of illumination. The detects Fourier components that have become unreliable estimators shift due aliasing, removes them from shift-estimate computation. In presence average precision is few hundredths pixel. Experimental data presented here show yields superior in aliasing when compared several...

Digital images can be forged easily with today's widely available image processing software. In this paper, we describe a passive approach to detect digital forgeries by checking inconsistencies of blocking artifact. Given image, find that the artifacts introduced during JPEG compression could used as "natural authentication code". A artifact measure is then proposed based on estimated quantization table using power spectrum DCT coefficient histogram. Experimental results also demonstrate...

Recently, Halevi et al. (CCS '11) proposed a cryptographic primitive called proofs of ownership (PoW) to enhance security client-side deduplication in cloud storage. In proof scheme, any owner the same file F can prove storage that he/she owns robust and efficient way, bounded leakage setting where certain amount efficiently-extractable information about is leaked. Following this work, we propose secure with following advantages: our scheme protects data confidentiality (and some partial...

Learning-based systems have been shown to be vulnerable evasion through adversarial data manipulation. These attacks studied under assumptions that the adversary has certain knowledge of either target model internals, its training dataset or at least classification scores it assigns input samples. In this paper, we investigate a much more constrained and realistic attack scenario wherein classifier is minimally exposed adversary, revealing on final decision (e.g., reject accept an sample)....

Searchable encryption (SE) provides a privacy-preserving mechanism for data users to search over encrypted stored on remote server. Researchers have designed number of SE schemes with high efficiency yet allowing some degree leakage profile the The leakage, however, should be further measured allow us understand what types attacks an scheme would encounter. This paper considers passive that make inferences based prior knowledge and observations queries issued by users. is in contrast...

With the omnipresence of camera phone and digital display, capturing digitally displayed image with are getting widely practiced. In context watermarking, this brings forth issue screen-shooting robustness. The key to acquiring robustness is designing a good noise layer that could represent distortions in deep-learning-based watermarking framework. However, it very difficult quantitatively formulate distortion since process too complex. order design an effective for robustness, we propose...

In this paper, we proposed an automated system to perform a live memory forensic analysis for mobile phones. We investigated the dynamic behavior of phone's volatile memory, and is useful in real-time evidence acquisition communication based applications. Different scenarios with varying parameters were investigated. Our experimental results showed that outgoing messages (from phone) have higher persistency than incoming messages. our experiments, consistently achieved 100% rate For...

Proofs of Retrievability (POR) is a cryptographic formulation for remotely auditing the integrity files stored in cloud, without keeping copy original local storage. In POR scheme, user Alice backups her data file together with some authentication to potentially dishonest cloud storage server Bob. Later, can periodically and verify using data, retrieving back file. Besides security, performances communication, overhead computation are major considerations. Shacham Waters (Asiacrypt '08) gave...

With the introduction of advanced monitoring and communication technologies, legacy power system is evolving toward fully digitalized smart grid. The IEC 61850 standards have emerged as popular for automation grid systems. However, modernization has an unintended consequence increased vulnerability to cyber-attacks. Among emerging cyber threats, in this article we focus on false data injection attacks against generators compliant systems since subject not yet intensively studied literature....

The use of multiple InSAR images the same area for phase unwrapping is discussed. It shown that by combining information in two images, ambiguity interval angle can be considerably lengthened, which will facilitate unwrapping. Three methods are proposed and their error characteristics discussed.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

In biometric identification, a fingerprint is typically represented as set of minutiae which are 2D points. A method [4] to protect the template hides by adding random points (known chaff) into original point set. The chaff added one-by-one, constrained requirement that no two close each other, until it impossible add more or sufficient number have been added. Therefore, if consists s points, and total m, then brute-force attacker expected examine half m chooses possibilities find original....

This paper presents tagged-MapReduce, a general extension to MapReduce that supports secure computing with mixed-sensitivity data on hybrid clouds. Tagged-MapReduce augments each key-value pair in sensitivity tag. enables fine-grained dataflow control during execution prevent leakage as well supporting expressive security policies and complex computations. Security constraints for preventing impose restrictions computation storage/transfer, hence, we present scheduling strategies can exploit...

Neural networks are susceptible to data inference attacks such as the model inversion attack and membership attack, where attacker could infer reconstruction of a sample from confidence scores predicted by target classifier. In this paper, we propose unified approach, namely purification framework, defend attacks. It purifies score vectors classifier reducing their dispersion. The purifier can be further specialized in defending particular via adversarial learning. We evaluate our approach...

The COVID-19 pandemic and Taiwan’s strict prevention policies from May 2020 to April 2023 significantly altered the epidemiology of viral infections. This study analyzed pediatric influenza cases at Wei Gong Memorial Hospital January 2015 March 2024. Clinical features were compared among children hospitalized during Omicron BA.2.3.7 (April–July 2022), BA.5 (August–December JN.1 (2024) waves those with in 2024 2015–2016. Between 2024, there 2729 A (36.6%), 974 B (13.1%), 3752 (50.3%) cases,...

Fuzzy extractors are recently proposed error-tolerant cryptographic primitives that potentially useful to protect biometric templates. However, there challenges in adopting these primitives. Firstly, fuzzy require the data obtained during both enrollment and verification be same feature representation. for better performance on ROC, multiple high quality samples can enrollment, which result an asymmetric setting whereby stored different representations. Secondly, only concern about strength...

Network middleboxes perform deep packet inspection (DPI) to detect anomalies and suspicious activities in network traffic. However, increasingly these traffic are encrypted can no longer make sense of them. A recent proposal by Sherry et al. (SIGCOMM 2015), named BlindBox, enables the middlebox a privacy-preserving manner. BlindBox deploys garbled circuit generate rules for purpose inspecting directly. setup latency (which could be 97s on ruleset 3,000 as reported) overhead size incurred...

Deep learning-based digital watermarking frameworks have been widely studied recently. Most existing methods adopt an ``encoder-noise layer-decoder''-based architecture where the embedding and extraction processes are accomplished separately by encoder decoder. However, one potential drawback of such a framework is that decoder may not be well coupled, resulting in fact embed some redundant features into host image thus influencing invisibility robustness whole algorithm. To address this...

In this paper, we propose a novel blind-source separation method to extract fetal ECG from single-channel signal measured on the abdomen of mother. The is mixture ECG, maternal and noise. key idea project into higher dimensions, then use an assumption statistical independence between components separate them mixtures. This achieved by applying singular value decomposition (SVD) spectrogram, followed iterated application independent component analysis (ICA) principle components. SVD...