arXiv subjects

Yu-Hsuan Huang

Publications and source records attributed to Yu-Hsuan Huang.

5 recordsLinked to original sources

arXiv2025

Single Document Image Highlight Removal via A Large-Scale Real-World Dataset and A Location-Aware Network

Reflective documents often suffer from specular highlights under ambient lighting, severely hindering text readability and degrading overall visual quality. Although recent deep learning methods show promise in highlight removal, they remain suboptimal for document images, primarily due to the lack of dedicated datasets and tailored architectural designs. To tackle these challenges, we present DocHR14K, a large-scale real-world dataset comprising 14,902 high-resolution image pairs across six document categories and various lighting conditions. To the best of our knowledge, this is the first high-resolution dataset for document highlight removal that captures a wide range of real-world lighting conditions. Additionally, motivated by the observation that the residual map between highlighted and clean images naturally reveals the spatial structure of highlight regions, we propose a simple yet effective Highlight Location Prior (HLP) to estimate highlight masks without human annotations. Building on this prior, we present the Location-Aware Laplacian Pyramid Highlight Removal Network (L2HRNet), which effectively removes highlights by leveraging estimated priors and incorporates diffusion module to restore details. Extensive experiments demonstrate that DocHR14K improves highlight removal under diverse lighting conditions. Our L2HRNet achieves state-of-the-art performance across three benchmark datasets, including a 5.01\% increase in PSNR and a 13.17\% reduction in RMSE on DocHR14K.

cs.CV↗

arXiv2025

Future Sight and Tough Fights: Revolutionizing Sequential Recommendation with FENRec

Sequential recommendation (SR) systems predict user preferences by analyzing time-ordered interaction sequences. A common challenge for SR is data sparsity, as users typically interact with only a limited number of items. While contrastive learning has been employed in previous approaches to address the challenges, these methods often adopt binary labels, missing finer patterns and overlooking detailed information in subsequent behaviors of users. Additionally, they rely on random sampling to select negatives in contrastive learning, which may not yield sufficiently hard negatives during later training stages. In this paper, we propose Future data utilization with Enduring Negatives for contrastive learning in sequential Recommendation (FENRec). Our approach aims to leverage future data with time-dependent soft labels and generate enduring hard negatives from existing data, thereby enhancing the effectiveness in tackling data sparsity. Experiment results demonstrate our state-of-the-art performance across four benchmark datasets, with an average improvement of 6.16\% across all metrics.

cs.IR↗

arXiv2024

Isogeny-based Group Signatures and Accountable Ring Signatures in QROM

We present the first provably secure isogeny-based group signature (GS) and accountable ring signature (ARS) in the quantum random oracle model (QROM). We do so via introducing and constructing an intermediate primitive called the openable sigma protocol and demonstrating that any such protocol gives rise to a secure GS and ARS. Furthermore, QROM security is guaranteed if an additional perfect unique-response property (which is achieved via our tailored construction) is satisfied. Previous works by Beullens et al. (Eurocrypt 2022, Asiacrypt 2020) proposed isogeny-based GS and ARS with better efficiency but were only analyzed in the classical random oracle model (CROM). It is well-known that CROM security does not generally translate to QROM security; with the growing relevance of isogeny-based constructions in post-quantum cryptography, the current state of the art is unsatisfactory. Moreover, the aforementioned existing isogeny-based signatures were recently affected by the Fiat-Shamir with aborts (FSwA) flaw discovered by Barbosa et al. and Devevey et al. (CRYPTO 2023), leaving the provable security of isogeny-based signatures open to question once again. Our constructions are not only immune to the FSwA flaw but also provide stronger QROM security. As current QROM-secure ARS and GS schemes are mostly lattice-based, we offer a robust post-quantum alternative should lattice assumptions weaken.

cs.CR↗

arXiv2022

Adaptive versus Static Multi-oracle Algorithms, and Quantum Security of a Split-key PRF

In the first part of the paper, we show a generic compiler that transforms any oracle algorithm that can query multiple oracles adaptively, i.e., can decide on which oracle to query at what point dependent on previous oracle responses, into a static algorithm that fixes these choices at the beginning of the execution. Compared to naive ways of achieving this, our compiler controls the blow-up in query complexity for each oracle individually, and causes a very mild blow-up only. In the second part of the paper, we use our compiler to show the security of the very efficient hash-based split-key PRF proposed by Giacon, Heuer and Poettering (PKC 2018), in the quantum random-oracle model. Using a split-key PRF as the key-derivation function gives rise to a secure KEM combiner. Thus, our result shows that the hash-based construction of Giacon et al. can be safely used in the context of quantum attacks, for instance to combine a well-established but only classically-secure KEM with a candidate KEM that is believed to be quantum-secure. Our security proof for the split-key PRF crucially relies on our adaptive-to-static compiler, but we expect our compiler to be useful beyond this particular application. Indeed, we discuss a couple of other, known results from the literature that would have profitted from our compiler, in that these works had to go though serious complications in order to deal with adaptivity.

cs.CR↗

arXiv2021

On the Compressed-Oracle Technique, and Post-Quantum Security of Proofs of Sequential Work

We revisit the so-called compressed oracle technique, introduced by Zhandry for analyzing quantum algorithms in the quantum random oracle model (QROM). To start off with, we offer a concise exposition of the technique, which easily extends to the parallel-query QROM, where in each query-round the considered algorithm may make several queries to the QROM in parallel. This variant of the QROM allows for a more fine-grained query-complexity analysis. Our main technical contribution is a framework that simplifies the use of (the parallel-query generalization of) the compressed oracle technique for proving query complexity results. With our framework in place, whenever applicable, it is possible to prove quantum query complexity lower bounds by means of purely classical reasoning. More than that, for typical examples the crucial classical observations that give rise to the classical bounds are sufficient to conclude the corresponding quantum bounds. We demonstrate this on a few examples, recovering known results (like the optimality of parallel Grover), but also obtaining new results (like the optimality of parallel BHT collision search). Our main target is the hardness of finding a $q$-chain with fewer than $q$ parallel queries, i.e., a sequence $x_0, x_1,\ldots, x_q$ with $x_i = H(x_{i-1})$ for all $1 \leq i \leq q$. The above problem of finding a hash chain is of fundamental importance in the context of proofs of sequential work. Indeed, as a concrete cryptographic application of our techniques, we prove that the "Simple Proofs of Sequential Work" proposed by Cohen and Pietrzak remains secure against quantum attacks. Such an analysis is not simply a matter of plugging in our new bound; the entire protocol needs to be analyzed in the light of a quantum attack. Thanks to our framework, this can now be done with purely classical reasoning.

quant-ph↗