arXiv subjects

Jiahui Liu

Publications and source records attributed to Jiahui Liu.

63 records · Page 4Linked to original sources

arXiv2023

Few-shot Learning for Inference in Medical Imaging with Subspace Feature Representations

Unlike the field of visual scene recognition where tremendous advances have taken place due to the availability of very large datasets to train deep neural networks, inference from medical images is often hampered by the fact that only small amounts of data may be available. When working with very small dataset problems, of the order of a few hundred items of data, the power of deep learning may still be exploited by using a model pre-trained on natural images as a feature extractor and carrying out classic pattern recognition techniques in this feature space, the so-called few-shot learning problem. In regimes where the dimension of this feature space is comparable to or even larger than the number of items of data, dimensionality reduction is a necessity and is often achieved by principal component analysis, i.e., singular value decomposition (SVD). In this paper, noting the inappropriateness of using SVD for this setting, we usher in and explore two alternatives based on discriminant analysis and non-negative matrix factorization (NMF). Using 14 different datasets spanning $11$ distinct disease types, we demonstrate that discriminant subspaces at low dimensions achieve significant improvements over SVD-based subspaces and the original feature space. We also show that NMF at modest dimensions is a competitive alternative to SVD in this setting.

math.NA↗

arXiv2023

GO-LDA: Generalised Optimal Linear Discriminant Analysis

Linear discriminant analysis (LDA) has been a useful tool in pattern recognition and data analysis research and practice. While linearity of class boundaries cannot always be expected, nonlinear projections through pre-trained deep neural networks have served to map complex data onto feature spaces in which linear discrimination has served well. The solution to binary LDA is obtained by eigenvalue analysis of within-class and between-class scatter matrices. It is well known that the multiclass LDA is solved by an extension to the binary LDA, a generalised eigenvalue problem, from which the largest subspace that can be extracted is of dimension one lower than the number of classes in the given problem. In this paper, we show that, apart from the first of the discriminant directions, the generalised eigenanalysis solution to multiclass LDA does neither yield orthogonal discriminant directions nor maximise discrimination of projected data along them. Surprisingly, to the best of our knowledge, this has not been noted in decades of literature on LDA. To overcome this drawback, we present a derivation with a strict theoretical support for sequentially obtaining discriminant directions that are orthogonal to previously computed ones and maximise in each step the Fisher criterion. We show distributions of projections along these axes and demonstrate that discrimination of data projected onto these discriminant directions has optimal separation, which is much higher than those from the generalised eigenvectors of the multiclass LDA. Using a wide range of benchmark tasks, we present a comprehensive empirical demonstration that on a number of pattern recognition and classification problems, the optimal discriminant subspaces obtained by the proposed method, referred to as GO-LDA (Generalised Optimal LDA), can offer superior accuracy.

cs.CV↗

arXiv2023

Robust Extrinsic Self-Calibration of Camera and Solid State LiDAR

This letter proposes an extrinsic calibration approach for a pair of monocular camera and prism-spinning solid-state LiDAR. The unique characteristics of the point cloud measured resulting from the flower-like scanning pattern is first disclosed as the vacant points, a type of outlier between foreground target and background objects. Unlike existing method using only depth continuous measurements, we use depth discontinuous measurements to retain more valid features and efficiently remove vacant points. The larger number of detected 3D corners thus contain more robust a priori information than usual which, together with the 2D corners detected by overlapping cameras and constrained by the proposed circularity and rectangularity rules, produce accurate extrinsic estimates. The algorithm is evaluated with real field experiments adopting both qualitative and quantitative performance criteria, and found to be superior to existing algorithms. The code is available on GitHub.

cs.RO↗

arXiv2022

Another Round of Breaking and Making Quantum Money: How to Not Build It from Lattices, and More

Public verification of quantum money has been one of the central objects in quantum cryptography ever since Wiesner's pioneering idea of using quantum mechanics to construct banknotes against counterfeiting. So far, we do not know any publicly-verifiable quantum money scheme that is provably secure from standard assumptions. In this work, we provide both negative and positive results for publicly verifiable quantum money. **In the first part, we give a general theorem, showing that a certain natural class of quantum money schemes from lattices cannot be secure. We use this theorem to break the recent quantum money scheme of Khesin, Lu, and Shor. **In the second part, we propose a framework for building quantum money and quantum lightning we call invariant money which abstracts some of the ideas of quantum money from knots by Farhi et al.(ITCS'12). In addition to formalizing this framework, we provide concrete hard computational problems loosely inspired by classical knowledge-of-exponent assumptions, whose hardness would imply the security of quantum lightning, a strengthening of quantum money where not even the bank can duplicate banknotes. **We discuss potential instantiations of our framework, including an oracle construction using cryptographic group actions and instantiations from rerandomizable functional encryption, isogenies over elliptic curves, and knots.

cs.CR↗

arXiv2022

Hidden Cosets and Applications to Unclonable Cryptography

In this work, we study a generalization of hidden subspace states to hidden coset states (first introduced by Aaronson and Christiano [STOC '12]). This notion was considered independently by Vidick and Zhang [Eurocrypt '21], in the context of proofs of quantum knowledge from quantum money schemes. We explore unclonable properties of coset states and several applications: - We show that assuming indistinguishability obfuscation (iO), hidden coset states possess a certain direct product hardness property, which immediately implies a tokenized signature scheme in the plain model. Previously, it was known only relative to an oracle, from a work of Ben-David and Sattath [QCrypt '17]. - Combining a tokenized signature scheme with extractable witness encryption, we give a construction of an unclonable decryption scheme in the plain model. The latter primitive was recently proposed by Georgiou and Zhandry [ePrint '20], who gave a construction relative to a classical oracle. - We conjecture that coset states satisfy a certain natural (information-theoretic) monogamy-of-entanglement property. Assuming this conjecture is true, we remove the requirement for extractable witness encryption in our unclonable decryption construction, by relying instead on compute-and-compare obfuscation for the class of unpredictable distributions. This conjecture was later proved by Culf and Vidick in a follow-up work. - Finally, we give a construction of a copy-protection scheme for pseudorandom functions (PRFs) in the plain model. Our scheme is secure either assuming iO, OWF, and extractable witness encryption, or assuming iO, OWF, compute-and-compare obfuscation for the class of unpredictable distributions, and the conjectured monogamy property mentioned above. This is the first example of a copy-protection scheme with provable security in the plain model for a class of functions that is not evasive.

cs.CR↗

arXiv2022

GPUMD: A package for constructing accurate machine-learned potentials and performing highly efficient atomistic simulations

We present our latest advancements of machine-learned potentials (MLPs) based on the neuroevolution potential (NEP) framework introduced in [Fan et al., Phys. Rev. B 104, 104309 (2021)] and their implementation in the open-source package GPUMD. We increase the accuracy of NEP models both by improving the radial functions in the atomic-environment descriptor using a linear combination of Chebyshev basis functions and by extending the angular descriptor with some four-body and five-body contributions as in the atomic cluster expansion approach. We also detail our efficient implementation of the NEP approach in graphics processing units as well as our workflow for the construction of NEP models, and we demonstrate their application in large-scale atomistic simulations. By comparing to state-of-the-art MLPs, we show that the NEP approach not only achieves above-average accuracy but also is far more computationally efficient. These results demonstrate that the GPUMD package is a promising tool for solving challenging problems requiring highly accurate, large-scale atomistic simulations. To enable the construction of MLPs using a minimal training set, we propose an active-learning scheme based on the latent space of a pre-trained NEP model. Finally, we introduce three separate Python packages, GPYUMD, CALORINE, and PYNEP, which enable the integration of GPUMD into Python workflows.

physics.comp-ph↗

arXiv2022

Beating Classical Impossibility of Position Verification

Chandran et al. (SIAM J. Comput.'14) formally introduced the cryptographic task of position verification, where they also showed that it cannot be achieved by classical protocols. In this work, we initiate the study of position verification protocols with classical verifiers. We identify that proofs of quantumness (and thus computational assumptions) are necessary for such position verification protocols. For the other direction, we adapt the proof of quantumness protocol by Brakerski et al. (FOCS'18) to instantiate such a position verification protocol. As a result, we achieve classically verifiable position verification assuming the quantum hardness of Learning with Errors. Along the way, we develop the notion of 1-of-2 non-local soundness for a natural non-local game for 1-of-2 puzzles, first introduced by Radian and Sattath (AFT'19), which can be viewed as a computational unclonability property. We show that 1-of-2 non-local soundness follows from the standard 2-of-2 soundness (and therefore the adaptive hardcore bit property), which could be of independent interest.

quant-ph↗

arXiv2020

New Approaches for Quantum Copy-Protection

Quantum copy protection uses the unclonability of quantum states to construct quantum software that provably cannot be pirated. Copy protection would be immensely useful, but unfortunately little is known about how to achieve it in general. In this work, we make progress on this goal, by giving the following results: - We show how to copy protect any program that cannot be learned from its input/output behavior, relative to a classical oracle. This improves on Aaronson [CCC'09], which achieves the same relative to a quantum oracle. By instantiating the oracle with post-quantum candidate obfuscation schemes, we obtain a heuristic construction of copy protection. -We show, roughly, that any program which can be watermarked can be copy detected, a weaker version of copy protection that does not prevent copying, but guarantees that any copying can be detected. Our scheme relies on the security of the assumed watermarking, plus the assumed existence of public key quantum money. Our construction is general, applicable to many recent watermarking schemes.

cs.CR↗

arXiv2017

Mining Functional Modules by Multiview-NMF of Phenome-Genome Association

Background: Mining gene modules from genomic data is an important step to detect gene members of pathways or other relations such as protein-protein interactions. In this work, we explore the plausibility of detecting gene modules by factorizing gene-phenotype associations from a phenotype ontology rather than the conventionally used gene expression data. In particular, the hierarchical structure of ontology has not been sufficiently utilized in clustering genes while functionally related genes are consistently associated with phenotypes on the same path in the phenotype ontology. Results: We propose a hierarchal Nonnegative Matrix Factorization (NMF)-based method, called Consistent Multiple Nonnegative Matrix Factorization (CMNMF), to factorize genome-phenome association matrix at two levels of the hierarchical structure in phenotype ontology for mining gene functional modules. CMNMF constrains the gene clusters from the association matrices at two consecutive levels to be consistent since the genes are annotated with both the child phenotype and the parent phenotype in the consecutive levels. CMNMF also restricts the identified phenotype clusters to be densely connected in the phenotype ontology hierarchy. In the experiments on mining functionally related genes from mouse phenotype ontology and human phenotype ontology, CMNMF effectively improved clustering performance over the baseline methods. Gene ontology enrichment analysis was also conducted to reveal interesting gene modules. Conclusions: Utilizing the information in the hierarchical structure of phenotype ontology, CMNMF can identify functional gene modules with more biological significance than the conventional methods. CMNMF could also be a better tool for predicting members of gene pathways and protein-protein interactions. Availability: https://github.com/nkiip/CMNMF

cs.LG↗