Searcharxiv⌕ Search

arXiv subjects

Lin Sun

Publications and source records attributed to Lin Sun.

At least 73 records · Page 4Linked to original sources
arXiv2024

CLIPer: Hierarchically Improving Spatial Representation of CLIP for Open-Vocabulary Semantic Segmentation

Contrastive Language-Image Pre-training (CLIP) exhibits strong zero-shot classification ability on various image-level tasks, leading to the research to adapt CLIP for pixel-level open-vocabulary semantic segmentation without additional training. The key is to improve spatial representation of image-level CLIP, such as replacing self-attention map at last layer with self-self attention map or vision foundation model based attention map. In this paper, we present a novel hierarchical framework, named CLIPer, that hierarchically improves spatial representation of CLIP. The proposed CLIPer includes an early-layer fusion module and a fine-grained compensation module. We observe that, the embeddings and attention maps at early layers can preserve spatial structural information. Inspired by this, we design the early-layer fusion module to generate segmentation map with better spatial coherence. Afterwards, we employ a fine-grained compensation module to compensate the local details using the self-attention maps of diffusion model. We conduct the experiments on seven segmentation datasets. Our proposed CLIPer achieves the state-of-the-art performance on these datasets. For instance, using ViT-L, CLIPer has the mIoU of 69.8% and 43.3% on VOC and COCO Object, outperforming ProxyCLIP by 9.2% and 4.1% respectively.

cs.CV↗
arXiv2024

On Bott--Samelson rings for Coxeter groups

We study the cohomology ring of the Bott--Samelson variety. We compute an explicit presentation of this ring via Soergel's result, which implies that it is a purely combinatorial invariant. We use the presentation to introduce the Bott--Samelson ring associated with a word in arbitrary Coxeter system by generators and relations. In general, it is a split quadratic complete intersection algebra with a triangular pattern of relations. By a result of Tate, it follows that it is a Koszul algebra and we provide a quadratic (reduced) Gr{ö}bner basis. Furthermore, we prove that it satisfies the whole Kähler package, including the Poincaré duality, the hard Lefschetz theorem, and the Hodge--Riemann bilinear relations.

math.RA↗
arXiv2024

Aegis:An Advanced LLM-Based Multi-Agent for Intelligent Functional Safety Engineering

Functional safety is a critical aspect of automotive engineering, encompassing all phases of a vehicle's lifecycle, including design, development, production, operation, and decommissioning. This domain involves highly knowledge-intensive tasks. This paper introduces Aegis: An Advanced LLM-Based Multi-Agent for Intelligent Functional Safety Engineering. Aegis is specifically designed to support complex functional safety tasks within the automotive sector. It is tailored to perform Hazard Analysis and Risk Assessment(HARA), document Functional Safety Requirements(FSR), and plan test cases for Automatic Emergency Braking(AEB) systems. The most advanced version, Aegis-Max, leverages Retrieval-Augmented Generation(RAG) and reflective mechanisms to enhance its capability in managing complex, knowledge-intensive tasks. Additionally, targeted prompt refinement by professional functional safety practitioners can significantly optimize Aegis's performance in the functional safety domain. This paper demonstrates the potential of Aegis to improve the efficiency and effectiveness of functional safety processes in automotive engineering.

cs.MA↗
arXiv2024

iSeg: An Iterative Refinement-based Framework for Training-free Segmentation

Stable diffusion has demonstrated strong image synthesis ability to given text descriptions, suggesting it to contain strong semantic clue for grouping objects. The researchers have explored employing stable diffusion for training-free segmentation. Most existing approaches refine cross-attention map by self-attention map once, demonstrating that self-attention map contains useful semantic information to improve segmentation. To fully utilize self-attention map, we present a deep experimental analysis on iteratively refining cross-attention map with self-attention map, and propose an effective iterative refinement framework for training-free segmentation, named iSeg. The proposed iSeg introduces an entropy-reduced self-attention module that utilizes a gradient descent scheme to reduce the entropy of self-attention map, thereby suppressing the weak responses corresponding to irrelevant global information. Leveraging the entropy-reduced self-attention module, our iSeg stably improves refined cross-attention map with iterative refinement. Further, we design a category-enhanced cross-attention module to generate accurate cross-attention map, providing a better initial input for iterative refinement. Extensive experiments across different datasets and diverse segmentation tasks reveal the merits of proposed contributions, leading to promising performance on diverse segmentation tasks. For unsupervised semantic segmentation on Cityscapes, our iSeg achieves an absolute gain of 3.8% in terms of mIoU compared to the best existing training-free approach in literature. Moreover, our proposed iSeg can support segmentation with different kinds of images and interactions. The project is available at https://linsun449.github.io/iSeg.

cs.CV↗
arXiv2024

Large $ Y_{3,2} $-tilings in 3-uniform hypergraphs

Let $Y_{3,2}$ be the 3-graph with two edges intersecting in two vertices. We prove that every 3-graph $ H $ on $ n $ vertices with at least $ \max \left \{ \binom{4αn}{3}, \binom{n}{3}-\binom{n-αn}{3} \right \}+o(n^3) $ edges contains a $Y_{3,2}$-tiling covering more than $ 4αn$ vertices, for sufficiently large $ n $ and $0<α< 1/4$. The bound on the number of edges is asymptotically best possible and solves a conjecture of the authors for 3-graphs that generalizes the Matching Conjecture of Erdős.

math.CO↗
arXiv2024

Enhancing Trust and Privacy in Distributed Networks: A Comprehensive Survey on Blockchain-based Federated Learning

While centralized servers pose a risk of being a single point of failure, decentralized approaches like blockchain offer a compelling solution by implementing a consensus mechanism among multiple entities. Merging distributed computing with cryptographic techniques, decentralized technologies introduce a novel computing paradigm. Blockchain ensures secure, transparent, and tamper-proof data management by validating and recording transactions via consensus across network nodes. Federated Learning (FL), as a distributed machine learning framework, enables participants to collaboratively train models while safeguarding data privacy by avoiding direct raw data exchange. Despite the growing interest in decentralized methods, their application in FL remains underexplored. This paper presents a thorough investigation into Blockchain-based FL (BCFL), spotlighting the synergy between blockchain's security features and FL's privacy-preserving model training capabilities. First, we present the taxonomy of BCFL from three aspects, including decentralized, separate networks, and reputation-based architectures. Then, we summarize the general architecture of BCFL systems, providing a comprehensive perspective on FL architectures informed by blockchain. Afterward, we analyze the application of BCFL in healthcare, IoT, and other privacy-sensitive areas. Finally, we identify future research directions of BCFL.

cs.CR↗
arXiv2024

Flat band effects on the ground-state BCS-BEC crossover in atomic Fermi gases in a quasi-two-dimensional Lieb lattice

The ground-state superfluid behavior of ultracold atomic Fermi gases with a short-range attractive interaction in a quasi-two-dimensional Lieb lattice is studied using BCS mean-field theory, within the context of BCS-BEC crossover. We find that the flat band leads to nontrivial exotic effects. As the Fermi level enters the flat band, both the pairing gap and the in-plane superfluid density exhibit an unusual power law as a function of interaction, with strongly enhanced quantum geometric effects, in addition to a dramatic increase of compressibility as the interaction approaches the BCS limit. As the Fermi level crosses the van Hove singularities, the character of pairing changes from particle-like to hole-like or vice versa. We present the computed phase diagram, in which a pair density wave state emerges at high densities with relatively strong interaction strength.

cond-mat.quant-gas↗
arXiv2024

PathAsst: A Generative Foundation AI Assistant Towards Artificial General Intelligence of Pathology

As advances in large language models (LLMs) and multimodal techniques continue to mature, the development of general-purpose multimodal large language models (MLLMs) has surged, offering significant applications in interpreting natural images. However, the field of pathology has largely remained untapped, particularly in gathering high-quality data and designing comprehensive model frameworks. To bridge the gap in pathology MLLMs, we present PathAsst, a multimodal generative foundation AI assistant to revolutionize diagnostic and predictive analytics in pathology. The development of PathAsst involves three pivotal steps: data acquisition, CLIP model adaptation, and the training of PathAsst's multimodal generative capabilities. Firstly, we collect over 207K high-quality pathology image-text pairs from authoritative sources. Leveraging the advanced power of ChatGPT, we generate over 180K instruction-following samples. Furthermore, we devise additional instruction-following data specifically tailored for invoking eight pathology-specific sub-models we prepared, allowing the PathAsst to effectively collaborate with these models, enhancing its diagnostic ability. Secondly, by leveraging the collected data, we construct PathCLIP, a pathology-dedicated CLIP, to enhance PathAsst's capabilities in interpreting pathology images. Finally, we integrate PathCLIP with the Vicuna-13b and utilize pathology-specific instruction-tuning data to enhance the multimodal generation capacity of PathAsst and bolster its synergistic interactions with sub-models. The experimental results of PathAsst show the potential of harnessing AI-powered generative foundation model to improve pathology diagnosis and treatment processes.

cs.CV↗
arXiv2024

BiSwift: Bandwidth Orchestrator for Multi-Stream Video Analytics on Edge

High-definition (HD) cameras for surveillance and road traffic have experienced tremendous growth, demanding intensive computation resources for real-time analytics. Recently, offloading frames from the front-end device to the back-end edge server has shown great promise. In multi-stream competitive environments, efficient bandwidth management and proper scheduling are crucial to ensure both high inference accuracy and high throughput. To achieve this goal, we propose BiSwift, a bi-level framework that scales the concurrent real-time video analytics by a novel adaptive hybrid codec integrated with multi-level pipelines, and a global bandwidth controller for multiple video streams. The lower-level front-back-end collaborative mechanism (called adaptive hybrid codec) locally optimizes the accuracy and accelerates end-to-end video analytics for a single stream. The upper-level scheduler aims to accuracy fairness among multiple streams via the global bandwidth controller. The evaluation of BiSwift shows that BiSwift is able to real-time object detection on 9 streams with an edge device only equipped with an NVIDIA RTX3070 (8G) GPU. BiSwift improves 10%$\sim$21% accuracy and presents 1.2$\sim$9$\times$ throughput compared with the state-of-the-art video analytics pipelines.

cs.NI↗
arXiv2024

UMIE: Unified Multimodal Information Extraction with Instruction Tuning

Multimodal information extraction (MIE) gains significant attention as the popularity of multimedia content increases. However, current MIE methods often resort to using task-specific model structures, which results in limited generalizability across tasks and underutilizes shared knowledge across MIE tasks. To address these issues, we propose UMIE, a unified multimodal information extractor to unify three MIE tasks as a generation problem using instruction tuning, being able to effectively extract both textual and visual mentions. Extensive experiments show that our single UMIE outperforms various state-of-the-art (SoTA) methods across six MIE datasets on three tasks. Furthermore, in-depth analysis demonstrates UMIE's strong generalization in the zero-shot setting, robustness to instruction variants, and interpretability. Our research serves as an initial step towards a unified MIE model and initiates the exploration into both instruction tuning and large language models within the MIE domain. Our code, data, and model are available at https://github.com/ZUCC-AI/UMIE

cs.AI↗
arXiv2023

The threshold for powers of tight Hamilton cycles in random hypergraphs

We investigate the occurrence of powers of tight Hamilton cycles in random hypergraphs. For every $r\ge 3$ and $k\ge 1$, we show that there exists a constant $C > 0$ such that if $p=p(n) \ge Cn^{-1/\binom{k+r-2}{r-1}}$ then asymptotically almost surely the random hypergraph $H^{(r)}(n,p)$ contains the $k$th power of a tight Hamilton cycle. This improves on a result of Parczyk and Person, who proved the same result under the assumption $p=ω\left(n^{-1/\binom{k+r-2}{r-1}}\right)$ using a second moment argument.

math.CO↗
arXiv2023

Competition of electronic correlation and reconstruction in La1-xSrxTiO3/SrTiO3 heterostructures

Electronic correlation and reconstruction are two important factors that play a critical role in shaping the magnetic and electronic properties of correlated low-dimensional systems. Here, we report a competition between the electronic correlation and structural reconstruction in La1-xSrxTiO3/SrTiO3 heterostructures by modulating material polarity and interfacial strain, respectively. The heterostructures exhibit a critical thickness (tc) at which a metal-to-insulator transition (MIT) abruptly occurs at certain thickness, accompanied by the coexistence of two- and three-dimensional (2D and 3D) carriers. Intriguingly, the tc exhibits a V-shaped dependence on the doping concentration of Sr, with the smallest tc value at x = 0.5. We attribute this V-shaped dependence to the competition between the electronic reconstruction (modulated by the polarity) and the electronic correlation (modulated by strain), which are borne out by the experimental results, including strain-dependent electronic properties and the evolution of 2D and 3D carriers. Our findings underscore the significance of the interplay between electronic reconstruction and correlation in the realization and utilization of emergent electronic functionalities in low-dimensional correlated systems.

cond-mat.str-el↗
arXiv2023

DSSP: A Distributed, SLO-aware, Sensing-domain-privacy-Preserving Architecture for Sensing-as-a-Service

In this paper, we propose DSSP, a Distributed, SLO-aware, Sensing-domain-privacy-Preserving architecture for Sensing-as-a-Service (SaS). DSSP addresses four major limitations of the current SaS architecture. First, to improve sensing quality and enhance geographic coverage, DSSP allows Independent sensing Administrative Domains (IADs) to participate in sensing services, while preserving the autonomy of control and privacy for individual domains. Second, DSSP enables a marketplace in which a sensing data seller (i.e., an IAD) can sell its sensing data to more than one buyer (i.e., cloud service provider (CSP)), rather than being locked in with just one CSP. Third, DSSP enables per-query tail-latency service-level-objective (SLO) guaranteed SaS. Fourth, DSSP enables distributed, rather than centralized, query scheduling, making SaS highly scalable. At the core of DSSP is the design of a budget decomposition technique that translates: (a) a query tail-latency SLO into exact task response time budgets for sensing tasks of the query dispatched to individual IADs; and (b) the task budget for a task arrived at an IAD into exact subtask queuing deadlines for subtasks of the task dispatched to individual edge nodes in each IAD. This enables IADs to allocate their internal resources independently and accurately to meet the task budgets and hence, query tail-latency SLO, based on a simple subtask-budget-aware earliest-deadline-first queuing (EDFQ) policy for all the subtasks. The performance and scalability of DSSP are evaluated and verified by both on-campus testbed experiment at small scale and simulation at large scale.

cs.DC↗
arXiv2023

The uniform supertrees with the extremal spectral radius

For a $hypergraph$ $\mathcal{G}=(V, E)$ consisting of a nonempty vertex set $V=V(\mathcal{G})$ and an edge set $E=E(\mathcal{G})$, its $adjacency$ $matrix$ $\mathcal {A}_{\mathcal{G}}=[(\mathcal {A}_{\mathcal{G}})_{ij}]$ is defined as $(\mathcal {A}_{\mathcal{G}})_{ij}=\sum_{e\in E_{ij}}\frac{1}{|e| - 1}$, where $E_{ij} = \{e \in E : i, j \in e\}$. The $spectral$ $radius$ of a hypergraph $\mathcal{G}$, denoted by $ρ(\mathcal {G})$, is the maximum modulus among all eigenvalues of $\mathcal {A}_{\mathcal{G}}$. In this paper, among all $k$-uniform ($k\geq 3$) supertrees with fixed number of vertices, the supertrees with the maximum, the second maximum and the minimum spectral radius are completely determined, respectively.

math.CO↗
arXiv2023

The unicyclic hypergraph with extremal spectral radius

For a $hypergraph$ $\mathcal{G}=(V, E)$ consisting of a nonempty vertex set $V=V(\mathcal{G})$ and an edge set $E=E(\mathcal{G})$, its $adjacency$ $matrix$ $\mathcal {A}_{\mathcal{G}}=[(\mathcal {A}_{\mathcal{G}})_{ij}]$ is defined as $(\mathcal {A}_{\mathcal{G}})_{ij}=\sum_{e\in E_{ij}}\frac{1}{|e| - 1}$, where $E_{ij} = \{e \in E \, |\, i, j \in e\}$.The $spectral$ $radius$ of a hypergraph $\mathcal{G}$, denoted by $ρ(\mathcal {G})$, is the maximum modulus among all eigenvalues of $\mathcal {A}_{\mathcal{G}}$. In this paper, among all $k$-uniform ($k\geq 3$) unicyclic hypergraphs with fixed number of vertices, the hypergraphs with the maximum and the second the maximum spectral radius are completely determined, respectively.

math.CO↗
arXiv2023

PAGE: A Position-Aware Graph-Based Model for Emotion Cause Entailment in Conversation

Conversational Causal Emotion Entailment (C2E2) is a task that aims at recognizing the causes corresponding to a target emotion in a conversation. The order of utterances in the conversation affects the causal inference. However, most current position encoding strategies ignore the order relation among utterances and speakers. To address the issue, we devise a novel position-aware graph to encode the entire conversation, fully modeling causal relations among utterances. The comprehensive experiments show that our method consistently achieves state-of-the-art performance on two challenging test sets, proving the effectiveness of our model. Our source code is available on Github: https://github.com/XiaojieGu/PAGE.

cs.CL↗
arXiv2022

An integer grid bridge sampler for the Bayesian inference of incomplete birth-death records

A one-to-one correspondence is established between the bridge path space of birth-death processes and the exclusive union of the product spaces of simplexes and integer grids. Formulae are derived for the exact counting of the integer grid bridges with fixed number of upward jumps. Then a uniform sampler over such restricted bridge path space is constructed. This leads to a Monte Carlo scheme, the integer grid bridge sampler, IGBS, to evaluate the transition probabilities of birth-death processes. Even the near zero probability of rare event could now be evaluated with controlled relative error. The IGBS based Bayesian inference for the incomplete birth-death observations is readily performed in demonstrating examples and in the analysis of a severely incomplete data set recording a real epidemic event. Comparison is performed with the basic bootstrap filter, an elementary sequential importance resampling algorithm. The haunting filtering failure has found no position in the new scheme.

stat.CO↗