arXiv subjects

Xiaojun Cheng

Publications and source records attributed to Xiaojun Cheng.

9 recordsLinked to original sources

arXiv2026

Unified Dominance Graph for Interval-Predicate Approximate Nearest Neighbor Search

Approximate Nearest Neighbor Search (ANNS) is a core primitive for unstructured data retrieval. Real-world applications--such as temporal databases, financial data analysis, and retrieval-augmented generation--often require hybrid queries whose valid objects are constrained by continuous interval attributes, such as lifespans or price ranges. We study Interval-Predicate ANNS (IPANNS), where validity is determined by a predicate between an object interval and a query interval. Existing range-filtering ANNS (RFANNS) methods are designed for single-dimensional scalar filters, but interval predicates such as containment and overlap rely on two coupled endpoint constraints. Treating endpoints as independent scalar attributes can incur large intersection overhead, while containment-specific methods lack a generalized indexing abstraction. In this paper, we propose the Unified Dominance Graph (UDG), a graph-indexing framework for the closed two-bound conjunctive fragment of IPANNS. For a chosen interval predicate, UDG maps object and query endpoints into a normalized two-dimensional dominance space and builds a dominance-labeled graph over the transformed coordinates. Containment, overlap, and other supported endpoint-bound predicates therefore reuse the same construction and search algorithms after semantic mapping, while each UDG instance remains tied to its selected predicate. UDG compresses query-state-specific proximity graphs into one compact index. To improve graph search under restrictive interval filters, we add validity-preserving patch edges that provide routing choices when few objects remain valid. Extensive evaluations on standard benchmarks and real-world datasets show that UDG achieves stable query performance across multiple interval relations and workloads, significantly outperforming existing hybrid search baselines while maintaining low indexing overhead.

cs.DB↗

arXiv2025

Reading Between the Timelines: RAG for Answering Diachronic Questions

While Retrieval-Augmented Generation (RAG) excels at injecting static, factual knowledge into Large Language Models (LLMs), it exhibits a critical deficit in handling longitudinal queries that require tracking entities and phenomena across time. This blind spot arises because conventional, semantically-driven retrieval methods are not equipped to gather evidence that is both topically relevant and temporally coherent for a specified duration. We address this challenge by proposing a new framework that fundamentally redesigns the RAG pipeline to infuse temporal logic. Our methodology begins by disentangling a user's query into its core subject and its temporal window. It then employs a specialized retriever that calibrates semantic matching against temporal relevance, ensuring the collection of a contiguous evidence set that spans the entire queried period. To enable rigorous evaluation of this capability, we also introduce the Analytical Diachronic Question Answering Benchmark (ADQAB), a challenging evaluation suite grounded in a hybrid corpus of real and synthetic financial news. Empirical results on ADQAB show that our approach yields substantial gains in answer accuracy, surpassing standard RAG implementations by 13% to 27%. This work provides a validated pathway toward RAG systems capable of performing the nuanced, evolutionary analysis required for complex, real-world questions. The dataset and code for this study are publicly available at https://github.com/kwunhang/TA-RAG.

cs.CL↗

arXiv2024

Exploring the Meaningfulness of Nearest Neighbor Search in High-Dimensional Space

Dense high dimensional vectors are becoming increasingly vital in fields such as computer vision, machine learning, and large language models (LLMs), serving as standard representations for multimodal data. Now the dimensionality of these vector can exceed several thousands easily. Despite the nearest neighbor search (NNS) over these dense high dimensional vectors have been widely used for retrieval augmented generation (RAG) and many other applications, the effectiveness of NNS in such a high-dimensional space remains uncertain, given the possible challenge caused by the "curse of dimensionality." To address above question, in this paper, we conduct extensive NNS studies with different distance functions, such as $L_1$ distance, $L_2$ distance and angular-distance, across diverse embedding datasets, of varied types, dimensionality and modality. Our aim is to investigate factors influencing the meaningfulness of NNS. Our experiments reveal that high-dimensional text embeddings exhibit increased resilience as dimensionality rises to higher levels when compared to random vectors. This resilience suggests that text embeddings are less affected to the "curse of dimensionality," resulting in more meaningful NNS outcomes for practical use. Additionally, the choice of distance function has minimal impact on the relevance of NNS. Our study shows the effectiveness of the embedding-based data representation method and can offer opportunity for further optimization of dense vector-related applications.

cs.LG↗

arXiv2023

GlobalMatch: Registration of Forest Terrestrial Point Clouds by Global Matching of Relative Stem Positions

Registering point clouds of forest environments is an essential prerequisite for LiDAR applications in precision forestry. State-of-the-art methods for forest point cloud registration require the extraction of individual tree attributes, and they have an efficiency bottleneck when dealing with point clouds of real-world forests with dense trees. We propose an automatic, robust, and efficient method for the registration of forest point clouds. Our approach first locates tree stems from raw point clouds and then matches the stems based on their relative spatial relationship to determine the registration transformation. The algorithm requires no extra individual tree attributes and has quadratic complexity to the number of trees in the environment, allowing it to align point clouds of large forest environments. Extensive experiments on forest terrestrial point clouds have revealed that our method inherits the effectiveness and robustness of the stem-based registration strategy while exceedingly increasing its efficiency. Besides, we introduce a new benchmark dataset that complements the very few existing open datasets for the development and evaluation of registration methods for forest point clouds. The source code of our method and the dataset are available at https://github.com/zexinyang/GlobalMatch.

cs.CV↗

arXiv2018

Pseudospin-valley coupled edge states in a photonic topological insulator

Pseudo-spin and valley degrees of freedom (DOFs) engineered in photonic analogues of topological insulators (TI) provide potential approaches to optical encoding and robust signal transport. Here we observe a ballistic edge state whose spin-valley indices are locked to the direction of propagation along the interface between a valley photonic crystal and a metacrystal emulating the quantum spin Hall effect. We demonstrate the inhibition of inter-valley scattering at a Y-junction formed at the interfaces between photonic TIs carrying different spin-valley Chern numbers. These results open up the possibility of using the valley DOF to control the flow of optical signals in 2D structures.

physics.app-ph↗

arXiv2017

Single-Parameter Scaling and Maximum Entropy inside Disordered One-Dimensional Systems: Theory and Experiment

The single-parameter scaling hypothesis relating the average and variance of the logarithm of the conductance is a pillar of the theory of electronic transport. We use a maximum-entropy ansatz to explore the logarithm of the energy density, $\ln {\cal W}(x)$, at a depth $x$ into a random one-dimensional system. Single-parameter scaling would be the special case in which $x=L$ (the system length). We find the result, confirmed in microwave measurements and computer simulations, that the average of $\ln {\cal W}(x)$ is independent of $L$ and equal to $-x/\ell$, with $\ell$ the mean free path. At the beginning of the sample, ${\rm var}[\ln {\cal W}(x)]$ rises linearly with $x$ and is also independent of $L$, with a sublinear increase near the sample output. At $x=L$ we find a correction to the value of ${\rm var}[\ln T]$ predicted by single-parameter scaling.

cond-mat.dis-nn↗

arXiv2016

Impact of surface reflection on transmission eigenvalue statistics and energy distributions inside random media

The impact of surface reflection upon transmission through and energy distributions within random media has generally been described in terms of the boundary extrapolation lengths $z_b, z_b'$ at the input and output end of an open sample, which are the distance beyond the sample surfaces at which the energy density within the sample extrapolates to zeroThe importance of reflection at the sample boundaries plays a key role in the scaling of transmission. Here we consider the impact of surface reflection on the propagation of diffusive waves in terms of the modification of the distribution of transmission eigenvalues (DTE). We review our finding of a transition in the analytical form of the DTE at the point that the sample length equals $|z_b-z_b'|$. The highest transmission eigenvalue for stronger asymmetry in boundary reflection is strictly smaller than unity. The average transmission and profiles of energy density inside the sample can still be described in terms of the sample length, $L$, and the boundary extrapolation lengths on both sides of the sample, $z_b, z_b'$. For localized waves, we find the energy density profile within the sample is a segment of the distribution that would be found in a longer sample with length $L+z_b+z_b'$. These results suggest new ways of controlling wave interference in both diffusive and localized systems by varying boundary reflectivity.

cond-mat.dis-nn↗

arXiv2014

Phase Singularity Diffusion

We follow the trajectories of phase singularities at nulls of intensity in the speckle pattern of waves transmitted through random media as the frequency of the incident radiation is scanned in microwave experiments and numerical simulations. Phase singularities are observed to diffuse with a linear increase of the square displacement with frequency shift. The product of the diffusion coefficient of phase singularities in the transmitted speckle pattern and the photon diffusion coefficient through the random medium is proportional to the square of the effective sample length. This provides the photon diffusion coefficient and a method for characterizing the motion of dynamic material systems.

cond-mat.dis-nn↗

arXiv2013

Transmission eigenvalues in random media with surface reflection

The impact of surface reflection on the statistics of transmission eigenvalues is a largely unexplored subject of fundamental and practical importance in statistical optics. Here, we develop a first-principles theory and confirm numerically that the distribution of transmission eigenvalues of diffusive waves exhibits a nonanalytic `transition' as the strength of surface reflection at one surface passes through a critical value while that at the other is fixed. Above the critical value, the highest transmission eigenvalue is strictly smaller than unity and decreases with increasing internal reflection. When the input and output surfaces are equally reflective, the highest transmission eigenvalue is unity and the transition disappears irrespective of the strength of surface reflection.

cond-mat.dis-nn↗