arXiv subjects

Dagang Li

Publications and source records attributed to Dagang Li.

14 recordsLinked to original sources

arXiv2026

CAP: A Scalable Benchmark for Evaluating Cross-Site Browser Agents with Complex Actions and Perception

Large language models are increasingly deployed as autonomous agents that interact with the web through browsers. While recent progress has been driven by benchmarks that evaluate end-to-end task success, these evaluations largely overlook two fundamental sources of difficulty in real web browsing: complex actions over rich user interfaces and visual perception of dynamically rendered content, especially in workflows that span multiple websites. We introduce CAP, a scalable benchmark for evaluating browser agents on cross-site, human-like web tasks that require non-trivial UI interactions and visual understanding. Specifically, we adopt a decomposition-and-recomposition pipeline that first abstracts each website into a structured site card capturing user-facing functions, complex execution operations, and perceptual requirements, and then recomposes these components into realistic cross-site workflows. Each task is therefore grounded in multiple specific operations on each website, enabling fine-grained diagnosis. Built on this framework, we construct 420 tasks across 108 real-world websites and 24 domains under careful quality control. Experiments on state-of-the-art browser agents using our verifiable agent-as-a-judge evaluation framework show low success rates and reveal that perception-heavy interactions remain a major bottleneck, exposing substantial gaps between current agents and real-world web browsing demands.

cs.AI↗

arXiv2026

Unification of Closed-Open Industrial Detection Scenarios: New Large-Scale Benchmarks,Challenges and Baselines

Large-scale Visual-Language Models (LVLMs) have achieved remarkable success in natural visual tasks, yet their application to industrial defect detection remains challenging due to two fundamental limitations: (i) the scarcity of large-scale industrial datasets that cover diverse defect categories across multiple domains, and (ii) the reliance on manual prompts (points, boxes, masks) that introduce subjective noise and lack text-visual interaction for fine-grained understanding. To address these challenges, we introduce a Large-Scale Multi-Modal Industrial Open-Closed benchmark (MMIOC-1M) containing over one million samples across $14$ super-categories, $29$ industrial scenes, and $351$ defect subcategories. To our knowledge, MMIOC-1M is the first unified largest benchmark supporting both open-vocabulary and closed-set industrial detection, providing valuable pre-training data for LVLMs in industrial scenarios. Furthermore, we propose a Refined Text-Visual Prompt Network (RTVPNet) that incorporates three key innovations: (1) an expert-assisted domain projection mechanism that enables rapid adaptation of general vision models to industrial domains, (2) an energy-based sparse sampling strategy that automatically generates refined visual prompts without manual intervention, and (3) a bidirectional text-visual interaction module that enhances cross-modal semantic alignment and understanding. Extensive experiments demonstrate that RTVPNet achieves state-of-the-art performance on MMIOC-1M, LVIS, and COCO benchmarks while maintaining computational efficiency. The dataset and code are available at https://github.com/hellozzk/MMIO.

cs.AI↗

arXiv2026

FMRFusion: Frequency-Aware Multi-View Representation Learning for Heterogeneous Image Fusion

Infrared and visible image fusion aims to generate a composite image that retains significant target information and preserves detailed textures, integrating two heterogeneous modalities. Previous image fusion methods typically adopt a single-module stacking approach to extract features from the two modalities. However, these approaches may result in incomplete learning of their distinct characteristics, thereby limiting the fusion effectiveness and constrain ing robustness in real-world heterogeneous data scenarios. To address these challenges, we propose FMRFusion, a frequency-aware multi-view representation learning network for Heterogeneous Image Fusion. A Multi-Scale Struc tural Perception Module is introduced to effectively capture discriminative structures, extracting fine-grained local structures and essential contextual information. A bilinear frequency decomposition mechanism is employed to sepa rate features into high-frequency and low-frequency components, enabling joint modeling of local details and global representations across different frequency domains. Moreover, a Cross-View Complementary Interaction is incorpo rated to explicitly model and fuse the complementary characteristics between reflected light information and radiative intensity responses, facilitating effective cross-view interaction. We further improve the Performance of the fused results by flow matching, which progressively refines the fused features by learning the transformation from coarse data to high-quality representations. Extensive experiments conducted on multiple benchmark datasets demonstrate that FMRFusion achieves superior and consistent performance across a range of fusion tasks, especially in nighttime scenarios

cs.CV↗

arXiv2026

FinBoardBench: Benchmarking Dynamic Wealth Management and Strategic Financial Reasoning of LLMs via Board Game Simulations

Recently, large language models (LLMs) have achieved superior performance in static financial reasoning and simple dynamic trading tasks. However, existing static financial benchmarks are insufficient to assess the dynamic wealth management and financial decision-making capabilities of LLMs in real-world environments. To bridge this gap, we present FinBoardBench, an evaluation suite based on three classic financial board games: Cashflow, Acquire, and Monopoly. FinBoardBench assesses a comprehensive set of financial skills, including personal cash flow management with debt balancing, corporate investment and acquisition forecasting, and competitive trade negotiations with asset auctions. Our experiments with 9 advanced LLMs reveal that while exhibiting basic long-term planning and investment logic, they fail to effectively leverage complex interactions for profit, and their strong static reasoning performance does not transform into successful dynamic decision-making. Notably, they tend to prioritize immediate asset acquisition over maintaining sufficient liquidity, making them vulnerable to financial crises triggered by random events. We hope that FinBoardBench can provide a valuable reference for more intelligent LLM-based decision-making systems in the future.

cs.CL↗

arXiv2026

Distilled Large Language Model-Driven Dynamic Sparse Expert Activation Mechanism

High inter-class similarity, extreme scale variation, and limited computational budgets hinder reliable visual recognition across diverse real-world data. Existing vision-centric and cross-modal approaches often rely on rigid fusion mechanisms and heavy annotation pipelines, leading to sub-optimal generalization. We propose the Distilled Large Language Model (LLM)-Driven Sparse Mixture-of-Experts (DS-MoE) framework, which integrates text-guided dynamic routing and lightweight multi-scale comprehension. The DS-MoE framework dynamically aligns textual semantics with defect-specific visual patterns through a sparse MoE architecture, where task-relevant experts are adaptively activated based on semantic relevance, resolving inter-class ambiguity. A lightweight MobileSAM encoder enables real-time inference while preserving multi-scale defect details. Extensive experiments on PCB, aluminum foil, and mold defect datasets demonstrate that our framework achieves superior performance compared to existing pure vision models. \textbf{DS-MoE} surpasses YOLOv8/YOLOX with gains of +13.9, +1.4, and +2.0 pp mAP@ 0.5:0.95 on BBMP, aluminum, and PCB, respectively, while also improving precision and recall.

cs.CV↗

arXiv2026

Can Large Language Models Resolve Semantic Discrepancy in Self-Destructive Subcultures? Evidence from Jirai Kei

Self-destructive behaviors are linked to complex psychological states and can be challenging to diagnose. These behaviors may be even harder to identify within subcultural groups due to their unique expressions. As large language models (LLMs) being deployed across various fields, some researchers have begun exploring their application for detecting self-destructive behaviors. Motivated by this, we investigate self-destructive behavior detection within subcultures using current LLM-based methods. However, these methods have two main challenges: (1) Knowledge Lag: Subcultural slang evolves rapidly, faster than LLMs' training cycles; and (2) Semantic Misalignment: it is challenging to grasp the specific and nuanced expressions unique to subcultures. To address these issues, we propose Subcultural Alignment Solver (SAS), a multi-agent framework that incorporates automatic retrieval and subculture alignment, significantly boosting the performance of LLMs in detecting self-destructive behavior. Our experimental results show that SAS outperforms the current advanced multi-agent framework OWL. Notably, it competes well with fine-tuned LLMs. We hope that SAS will advance the field of self-destructive behavior detection in subcultural contexts and serve as a valuable resource for future researchers.

cs.CL↗

arXiv2025

What Factors Affect LLMs and RLLMs in Financial Question Answering?

Recently, large language models (LLMs) and reasoning large language models (RLLMs) have gained considerable attention from many researchers. RLLMs enhance the reasoning capabilities of LLMs through Long Chain-of-Thought (Long CoT) processes, significantly improving the performance of LLMs in addressing complex problems. However, there are few works that systematically explore what methods can fully unlock the performance of LLMs and RLLMs within the financial domain. To investigate the impact of various methods on LLMs and RLLMs, we utilize five LLMs and four RLLMs to assess the effects of prompting methods, agentic frameworks, and multilingual alignment methods on financial question-answering tasks. Our research findings indicate: (1) Current prompting methods and agent frameworks enhance the performance of LLMs in financial question answering by simulating Long CoT; (2) RLLMs possess inherent Long CoT capabilities, which limits the effectiveness of conventional methods in further enhancing their performance; (3) Current advanced multilingual alignment methods primarily improve the multilingual performance of LLMs by extending the reasoning length, which yields minimal benefits for RLLMs. Additionally, we discuss strategies for enhancing the performance of LLMs and RLLMs in financial question answering, which may serve as a inspiration for future improvements. We hope that this study can serve as an important reference for LLMs and RLLMs in the field of financial question answering.

cs.CL↗

arXiv2025

LinkTo-Anime: A 2D Animation Optical Flow Dataset from 3D Model Rendering

Existing optical flow datasets focus primarily on real-world simulation or synthetic human motion, but few are tailored to Celluloid(cel) anime character motion: a domain with unique visual and motion characteristics. To bridge this gap and facilitate research in optical flow estimation and downstream tasks such as anime video generation and line drawing colorization, we introduce LinkTo-Anime, the first high-quality dataset specifically designed for cel anime character motion generated with 3D model rendering. LinkTo-Anime provides rich annotations including forward and backward optical flow, occlusion masks, and Mixamo Skeleton. The dataset comprises 395 video sequences, totally 24,230 training frames, 720 validation frames, and 4,320 test frames. Furthermore, a comprehensive benchmark is constructed with various optical flow estimation methods to analyze the shortcomings and limitations across multiple datasets.

cs.CV↗

arXiv2025

Zero-Shot Subject-Centric Generation for Creative Application Using Entropy Fusion

Generative models are widely used in visual content creation. However, current text-to-image models often face challenges in practical applications-such as textile pattern design and meme generation-due to the presence of unwanted elements that are difficult to separate with existing methods. Meanwhile, subject-reference generation has emerged as a key research trend, highlighting the need for techniques that can produce clean, high-quality subject images while effectively removing extraneous components. To address this challenge, we introduce a framework for reliable subject-centric image generation. In this work, we propose an entropy-based feature-weighted fusion method to merge the informative cross-attention features obtained from each sampling step of the pretrained text-to-image model FLUX, enabling a precise mask prediction and subject-centric generation. Additionally, we have developed an agent framework based on Large Language Models (LLMs) that translates users' casual inputs into more descriptive prompts, leading to highly detailed image generation. Simultaneously, the agents extract primary elements of prompts to guide the entropy-based feature fusion, ensuring focused primary element generation without extraneous components. Experimental results and user studies demonstrate our methods generates high-quality subject-centric images, outperform existing methods or other possible pipelines, highlighting the effectiveness of our approach.

cs.CV↗

arXiv2023

Modeling and Analysis on Efficiency Degradation of Lithium-ion Batteries

Efficiency of Battery Energy Storage Systems (BESSs) is increasingly critical as renewable energy generation becomes more prevalent on the grid. Therefore, it is necessary to study the energy efficiency of lithium-ion batteries, which are typically used in BESSs. The purpose of this study is to propose the State of Efficiency (SOE) as a measure of how efficiently batteries transfer energy, and to analyze what factors affect the SOE of a battery throughout its lifetime. Using NASA's data set, we measure the SOE of Nickel-Cobalt-Aluminum (NCA) lithium-ion batteries by calculating the ratio of energy generated and consumed during discharge and charge phases. A linear trend was observed in the SOE trajectories, which is confirmed by the Mann-Kendall (MK) trend test. Following that, a linear SOE degradation model was presented. Further analysis shows that ambient temperature, discharge current, and cutoff voltage all affect SOE in different ways. Using the SOE and its behavior observed in this study, Battery Management Systems (BMS) can improve the energy efficiency of batteries by adjusting operating conditions or developing better management strategies.

eess.SY↗

arXiv2020

Meta-Key: A Secure Data-Sharing Protocol under Blockchain-Based Decentralised Storage Architecture

In this letter we propose Meta-key, a data-sharing mechanism that enables users share their encrypted data under a blockchain-based decentralized storage architecture. All the data-encryption keys are encrypted by the owner's public key and put onto the blockchain for safe and secure storage and easy key-management. Encrypted data are stored in dedicated storage nodes and proxy re-encryption mechanism is used to ensure secure data-sharing in the untrusted environment. Security analysis of our model shows that the proxy re-encryption adopted in our system is naturally free from collusion-attack due to the specific architecture of Meta-key.

cs.DC↗

arXiv2019

Matrix Bloom Filter: An Efficient Probabilistic Data Structure for 2-tuple Batch Lookup

With the growing scale of big data, probabilistic structures receive increasing popularity for efficient approximate storage and query processing. For example, Bloom filters (BF) can achieve satisfactory performance for approximate membership existence query at the expense of false positives. However, a standard Bloom filter can only handle univariate data and single membership existence query, which is insufficient for OLAP and machine learning applications. In this paper, we focus on a common multivariate data type, namely, 2-tuples, or equivalently, key-value pairs. We design the matrix Bloom filter as a high-dimensional extension of the standard Bloom filter. This new probabilistic data structure can not only insert and lookup a single 2-tuple efficiently, but also support these operations efficiently in batches --- a key requirement for OLAP and machine learning tasks. To further balance the insertion and query efficiency for different workload patterns, we propose two variants, namely, the maximum adaptive matrix BF and minimum storage matrix BF. Through both theoretical and empirical studies, we show the performance of matrix Bloom filter is superior on datasets with common statistical distributions; and even without them, it just degrades to a standard Bloom filter.

cs.DS↗

arXiv2019

Flattened Exponential Histogram for Sliding Window Queries over Data Streams

The Basic Counting problem [1] is one of the most fundamental and critical streaming problems of sliding window queries over data streams. Given a stream of 0's and 1's, the purpose of this problem is to estimate the number of 1's in the last N elements (or time units) seen from the stream. Its solution can be used as building blocks to solve numerous more complex problems such as heavy hitter, frequency estimation, distinct counting, etc. In this paper, we present the flattened exponential histogram (FEH) model for the Basic Counting problem. Our model improves over the exponential histogram [1], [2], a well-received deterministic technique for Basic Counting problem, with respect to accuracy and memory utilization most of the time in practice. Extensive experimental results on real-world datasets show that with the same memory footprint, the accuracy of our model is between 4 to 15 and on average 7 times better than that of the exponential histogram, while the speed is roughly the same.

cs.DS↗

arXiv2017

Introduction of Improved Repairing Locality into Secret Sharing Schemes with Perfect Security

Repairing locality is an appreciated feature for distributed storage, in which a damaged or lost data share can be repaired by accessing a subset of other shares much smaller than is required for decoding the complete data. However for Secret Sharing (SS) schemes, it has been proven theoretically that local repairing can not be achieved with perfect security for the majority of threshold SS schemes, where all the shares are equally regarded in both secret recovering and share repairing. In this paper we make an attempt on decoupling the two processes to make secure local repairing possible. Dedicated repairing redundancies only for the repairing process are generated, which are random numbers to the original secret. Through this manner a threshold SS scheme with improved repairing locality is achieved on the condition that security of repairing redundancies is ensured, or else our scheme degenerates into a perfect access structure that is equivalent to the best existing schemes can do. To maximize security of the repairing redundancies, a random placement mechanism is also proposed.

cs.CR↗