arXiv subjects

Bing Li

Publications and source records attributed to Bing Li.

At least 163 records · Page 9Linked to original sources

arXiv2025

Rationale-guided Prompting for Knowledge-based Visual Question Answering

Recently, Large Language Models (LLMs) have been used for knowledge-based Visual Question Answering (VQA). Despite the encouraging results of previous studies, prior methods prompt LLMs to predict answers directly, neglecting intermediate thought processes. We argue that prior methods do not sufficiently activate the capacities of LLMs. We propose a framework called PLRH that Prompts LLMs with Rationale Heuristics for knowledge-based VQA. The PLRH prompts LLMs with Chain of Thought (CoT) to generate rationale heuristics, i.e., intermediate thought processes, and then leverages the rationale heuristics to inspire LLMs to predict answers. Experiments show that our approach outperforms the existing baselines by more than 2.2 and 2.1 on OK-VQA and A-OKVQA, respectively.

cs.CL↗

arXiv2025

Multi-Agents Based on Large Language Models for Knowledge-based Visual Question Answering

Large Language Models (LLMs) have achieved impressive results in knowledge-based Visual Question Answering (VQA). However existing methods still have challenges: the inability to use external tools autonomously, and the inability to work in teams. Humans tend to know whether they need to use external tools when they encounter a new question, e.g., they tend to be able to give a direct answer to a familiar question, whereas they tend to use tools such as search engines when they encounter an unfamiliar question. In addition, humans also tend to collaborate and discuss with others to get better answers. Inspired by this, we propose the multi-agent voting framework. We design three LLM-based agents that simulate different levels of staff in a team, and assign the available tools according to the levels. Each agent provides the corresponding answer, and finally all the answers provided by the agents are voted to get the final answer. Experiments on OK-VQA and A-OKVQA show that our approach outperforms other baselines by 2.2 and 1.0, respectively.

cs.CL↗

arXiv2025

LaMPE: Length-aware Multi-grained Positional Encoding for Adaptive Long-context Scaling Without Training

Large language models (LLMs) experience significant performance degradation when the input exceeds the pretraining context window, primarily due to the out-of-distribution (OOD) behavior of Rotary Position Embedding (RoPE). Recent studies mitigate this problem by remapping OOD positions into the in-distribution range with fixed mapping strategies, ignoring the dynamic relationship between input length and the model's effective context window. To this end, we propose Length-aware Multi-grained Positional Encoding (LaMPE), a training-free method that fully utilizes the model's effective context window for adaptive long-context scaling in LLMs. Motivated by the left-skewed frequency distribution of relative positions, LaMPE establishes a dynamic relationship between mapping length and input length through a parametric scaled sigmoid function to adaptively allocate positional capacity across varying input lengths. Meanwhile, LaMPE devises a novel multi-grained attention mechanism that strategically allocates positional resolution across different sequence regions to capture both fine-grained locality and long-range dependencies. Our method can be seamlessly applied to a wide range of RoPE-based LLMs without training. Extensive experiments on three representative LLMs across five mainstream long-context benchmarks demonstrate that LaMPE achieves significant performance improvements compared to existing length extrapolation methods. The code will be released at https://github.com/scar-on/LaMPE.

cs.CL↗

arXiv2025

CompressKV: Semantic Retrieval Heads Know What Tokens are Not Important Before Generation

Recent advances in large language models (LLMs) have significantly boosted long-context processing. However, the increasing key-value (KV) cache size poses critical challenges to memory and execution efficiency. Most KV cache compression methods rely on heuristic token eviction using all attention heads in Grouped Query Attention (GQA)-based LLMs. This method ignores the different functionalities of attention heads, leading to the eviction of critical tokens and thus degrades the performance of LLMs. To address the issue above, instead of using all the attention heads in GQA-based LLMs to determine important tokens as in the previous work, we first identify the attention heads in each layer that are not only capable of retrieving the initial and final tokens of a prompt, but also capable of retrieving important tokens within the text and attending to their surrounding semantic context. Afterwards, we exploit such heads to determine the important tokens and retain their corresponding KV cache pairs. Furthermore, we analyze the cache eviction error of each layer individually and introduce a layer-adaptive KV cache allocation strategy. Experimental results demonstrate the proposed CompressKV consistently outperforms state-of-the-art approaches under various memory budgets on LongBench and Needle-in-a-Haystack benchmarks. Our code is publicly available at: https://github.com/TUDa-HWAI/CompressKV.git.

cs.CL↗

arXiv2025

Point Cloud Self-supervised Learning via 3D to Multi-view Masked Learner

Recently, multi-modal masked autoencoders (MAE) has been introduced in 3D self-supervised learning, offering enhanced feature learning by leveraging both 2D and 3D data to capture richer cross-modal representations. However, these approaches have two limitations: (1) they inefficiently require both 2D and 3D modalities as inputs, even though the inherent multi-view properties of 3D point clouds already contain 2D modality. (2) input 2D modality causes the reconstruction learning to unnecessarily rely on visible 2D information, hindering 3D geometric representation learning. To address these challenges, we propose a 3D to Multi-View Learner (Multi-View ML) that only utilizes 3D modalities as inputs and effectively capture rich spatial information in 3D point clouds. Specifically, we first project 3D point clouds to multi-view 2D images at the feature level based on 3D-based pose. Then, we introduce two components: (1) a 3D to multi-view autoencoder that reconstructs point clouds and multi-view images from 3D and projected 2D features; (2) a multi-scale multi-head (MSMH) attention mechanism that facilitates local-global information interactions in each decoder transformer block through attention heads at various scales. Additionally, a novel two-stage self-training strategy is proposed to align 2D and 3D representations. Our method outperforms state-of-the-art counterparts across various downstream tasks, including 3D classification, part segmentation, and object detection.

cs.CV↗

arXiv2025

HLSTester: Efficient Testing of Behavioral Discrepancies with LLMs for High-Level Synthesis

In high-level synthesis (HLS), C/C++ programs with synthesis directives are used to generate circuits for FPGA implementations. However, hardware-specific and platform-dependent characteristics in these implementations can introduce behavioral discrepancies between the original C/C++ programs and the circuits after high-level synthesis. Existing methods for testing behavioral discrepancies in HLS are still immature, and the testing workflow requires significant human efforts. To address this challenge, we propose HLSTester, a large language model (LLM) aided testing framework that efficiently detects behavioral discrepancies in HLS. To mitigate hallucinations in LLMs and enhance prompt quality, the testbenches for original C/C++ programs are leveraged to guide LLMs in generating HLS-compatible testbenches, effectively eliminating certain traditional C/C++ constructs that are incompatible with HLS tools. Key variables are pinpointed through a backward slicing technique in both C/C++ and HLS programs to monitor their runtime spectra, enabling an in-depth analysis of the discrepancy symptoms. To reduce test time, a testing input generation mechanism is introduced to integrate dynamic mutation with insights from an LLM-based progressive reasoning chain. In addition, repetitive hardware testing is skipped by a redundancy-aware filtering technique for the generated test inputs. Experimental results demonstrate that the proposed LLM-aided testing framework significantly accelerates the testing workflow while achieving higher testbench simulation pass rates compared with the traditional method and the direct use of LLMs on the same HLS programs.

cs.SE↗

arXiv2025

EEvAct: Early Event-Based Action Recognition with High-Rate Two-Stream Spiking Neural Networks

Recognizing human activities early is crucial for the safety and responsiveness of human-robot and human-machine interfaces. Due to their high temporal resolution and low latency, event-based vision sensors are a perfect match for this early recognition demand. However, most existing processing approaches accumulate events to low-rate frames or space-time voxels which limits the early prediction capabilities. In contrast, spiking neural networks (SNNs) can process the events at a high-rate for early predictions, but most works still fall short on final accuracy. In this work, we introduce a high-rate two-stream SNN which closes this gap by outperforming previous work by 2% in final accuracy on the large-scale THU EACT-50 dataset. We benchmark the SNNs within a novel early event-based recognition framework by reporting Top-1 and Top-5 recognition scores for growing observation time. Finally, we exemplify the impact of these methods on a real-world task of early action triggering for human motion capture in sports.

cs.CV↗

arXiv2025

Doubly robust estimation of causal effects for random object outcomes with continuous treatments

Causal inference is central to statistics and scientific discovery, enabling researchers to identify cause-and-effect relationships beyond associations. While traditionally studied within Euclidean spaces, contemporary applications increasingly involve complex, non-Euclidean data structures that reside in abstract metric spaces, known as random objects, such as images, shapes, networks, and distributions. This paper introduces a novel framework for causal inference with continuous treatments applied to non-Euclidean data. To address the challenges posed by the lack of linear structures, we leverage Hilbert space embeddings of the metric spaces to facilitate Fréchet mean estimation and causal effect mapping. Motivated by a study on the impact of exposure to fine particulate matter on age-at-death distributions across U.S. counties, we propose a nonparametric, doubly-debiased causal inference approach for outcomes as random objects with continuous treatments. Our framework can accommodate moderately high-dimensional vector-valued confounders and derive efficient influence functions for estimation to ensure both robustness and interpretability. We establish rigorous asymptotic properties of the cross-fitted estimators and employ conformal inference techniques for counterfactual outcome prediction. Validated through numerical experiments and applied to real-world environmental data, our framework extends causal inference methodologies to complex data structures, broadening its applicability across scientific disciplines.

stat.ME↗

arXiv2025

Pay Less Attention to Deceptive Artifacts: Robust Detection of Compressed Deepfakes on Online Social Networks

With the rapid advancement of deep learning, particularly through generative adversarial networks (GANs) and diffusion models (DMs), AI-generated images, or ``deepfakes", have become nearly indistinguishable from real ones. These images are widely shared across Online Social Networks (OSNs), raising concerns about their misuse. Existing deepfake detection methods overlook the ``block effects" introduced by compression in OSNs, which obscure deepfake artifacts, and primarily focus on raw images, rarely encountered in real-world scenarios. To address these challenges, we propose PLADA (Pay Less Attention to Deceptive Artifacts), a novel framework designed to tackle the lack of paired data and the ineffective use of compressed images. PLADA consists of two core modules: Block Effect Eraser (B2E), which uses a dual-stage attention mechanism to handle block effects, and Open Data Aggregation (ODA), which processes both paired and unpaired data to improve detection. Extensive experiments across 26 datasets demonstrate that PLADA achieves a remarkable balance in deepfake detection, outperforming SoTA methods in detecting deepfakes on OSNs, even with limited paired data and compression. More importantly, this work introduces the ``block effect" as a critical factor in deepfake detection, providing a robust solution for open-world scenarios. Our code is available at https://github.com/ManyiLee/PLADA.

cs.CV↗

arXiv2025

Belted and Ensembled Neural Network for Linear and Nonlinear Sufficient Dimension Reduction

We introduce a unified, flexible, and easy-to-implement framework of sufficient dimension reduction that can accommodate both linear and nonlinear dimension reduction, and both the conditional distribution and the conditional mean as the targets of estimation. This unified framework is achieved by a specially structured neural network -- the Belted and Ensembled Neural Network (BENN) -- that consists of a narrow latent layer, which we call the belt, and a family of transformations of the response, which we call the ensemble. By strategically placing the belt at different layers of the neural network, we can achieve linear or nonlinear sufficient dimension reduction, and by choosing the appropriate transformation families, we can achieve dimension reduction for the conditional distribution or the conditional mean. Moreover, thanks to the advantage of the neural network, the method is very fast to compute, overcoming a computation bottleneck of the traditional sufficient dimension reduction estimators, which involves the inversion of a matrix of dimension either p or n. We develop the algorithm and convergence rate of our method, compare it with existing sufficient dimension reduction methods, and apply it to two data examples.

stat.ML↗

arXiv2025

Visual-Instructed Degradation Diffusion for All-in-One Image Restoration

Image restoration tasks like deblurring, denoising, and dehazing usually need distinct models for each degradation type, restricting their generalization in real-world scenarios with mixed or unknown degradations. In this work, we propose \textbf{Defusion}, a novel all-in-one image restoration framework that utilizes visual instruction-guided degradation diffusion. Unlike existing methods that rely on task-specific models or ambiguous text-based priors, Defusion constructs explicit \textbf{visual instructions} that align with the visual degradation patterns. These instructions are grounded by applying degradations to standardized visual elements, capturing intrinsic degradation features while agnostic to image semantics. Defusion then uses these visual instructions to guide a diffusion-based model that operates directly in the degradation space, where it reconstructs high-quality images by denoising the degradation effects with enhanced stability and generalizability. Comprehensive experiments demonstrate that Defusion outperforms state-of-the-art methods across diverse image restoration tasks, including complex and real-world degradations.

cs.CV↗

arXiv2025

Reversing Flow for Image Restoration

Image restoration aims to recover high-quality (HQ) images from degraded low-quality (LQ) ones by reversing the effects of degradation. Existing generative models for image restoration, including diffusion and score-based models, often treat the degradation process as a stochastic transformation, which introduces inefficiency and complexity. In this work, we propose ResFlow, a novel image restoration framework that models the degradation process as a deterministic path using continuous normalizing flows. ResFlow augments the degradation process with an auxiliary process that disambiguates the uncertainty in HQ prediction to enable reversible modeling of the degradation process. ResFlow adopts entropy-preserving flow paths and learns the augmented degradation flow by matching the velocity field. ResFlow significantly improves the performance and speed of image restoration, completing the task in fewer than four sampling steps. Extensive experiments demonstrate that ResFlow achieves state-of-the-art results across various image restoration benchmarks, offering a practical and efficient solution for real-world applications.

cs.CV↗

arXiv2025

IssueCourier: Multi-Relational Heterogeneous Temporal Graph Neural Network for Open-Source Issue Assignment

Issue assignment plays a critical role in open-source software (OSS) maintenance, which involves recommending the most suitable developers to address the reported issues. Given the high volume of issue reports in large-scale projects, manually assigning issues is tedious and costly. Previous studies have proposed automated issue assignment approaches that primarily focus on modeling issue report textual information, developers' expertise, or interactions between issues and developers based on historical issue-fixing records. However, these approaches often suffer from performance limitations due to the presence of incorrect and missing labels in OSS datasets, as well as the long tail of developer contributions and the changes of developer activity as the project evolves. To address these challenges, we propose IssueCourier, a novel Multi-Relational Heterogeneous Temporal Graph Neural Network approach for issue assignment. Specifically, we formalize five key relationships among issues, developers, and source code files to construct a heterogeneous graph. Then, we further adopt a temporal slicing technique that partitions the graph into a sequence of time-based subgraphs to learn stage-specific patterns. Furthermore, we provide a benchmark dataset with relabeled ground truth to address the problem of incorrect and missing labels in existing OSS datasets. Finally, to evaluate the performance of IssueCourier, we conduct extensive experiments on our benchmark dataset. The results show that IssueCourier can improve over the best baseline up to 45.49% in top-1 and 31.97% in MRR.

cs.SE↗

arXiv2025

Integer Binary-Range Alignment Neuron for Spiking Neural Networks

Spiking Neural Networks (SNNs) are noted for their brain-like computation and energy efficiency, but their performance lags behind Artificial Neural Networks (ANNs) in tasks like image classification and object detection due to the limited representational capacity. To address this, we propose a novel spiking neuron, Integer Binary-Range Alignment Leaky Integrate-and-Fire to exponentially expand the information expression capacity of spiking neurons with only a slight energy increase. This is achieved through Integer Binary Leaky Integrate-and-Fire and range alignment strategy. The Integer Binary Leaky Integrate-and-Fire allows integer value activation during training and maintains spike-driven dynamics with binary conversion expands virtual timesteps during inference. The range alignment strategy is designed to solve the spike activation limitation problem where neurons fail to activate high integer values. Experiments show our method outperforms previous SNNs, achieving 74.19% accuracy on ImageNet and 66.2% mAP@50 and 49.1% mAP@50:95 on COCO, surpassing previous bests with the same architecture by +3.45% and +1.6% and +1.8%, respectively. Notably, our SNNs match or exceed ANNs' performance with the same architecture, and the energy efficiency is improved by 6.3${\times}$.

cs.NE↗

arXiv2025

Rethinking Agent Design: From Top-Down Workflows to Bottom-Up Skill Evolution

Most LLM-based agent frameworks adopt a top-down philosophy: humans decompose tasks, define workflows, and assign agents to execute each step. While effective on benchmark-style tasks, such systems rely on designer updates and overlook agents' potential to learn from experience. Recently, Silver and Sutton(2025) envision a shift into a new era, where agents could progress from a stream of experiences. In this paper, we instantiate this vision of experience-driven learning by introducing a bottom-up agent paradigm that mirrors the human learning process. Agents acquire competence through a trial-and-reasoning mechanism-exploring, reflecting on outcomes, and abstracting skills over time. Once acquired, skills can be rapidly shared and extended, enabling continual evolution rather than static replication. As more agents are deployed, their diverse experiences accelerate this collective process, making bottom-up design especially suited for open-ended environments. We evaluate this paradigm in Slay the Spire and Civilization V, where agents perceive through raw visual inputs and act via mouse outputs, the same as human players. Using a unified, game-agnostic codebase without any game-specific prompts or privileged APIs, our bottom-up agents acquire skills entirely through autonomous interaction, demonstrating the potential of the bottom-up paradigm in complex, real-world environments. Our code is available at https://github.com/AngusDujw/Bottom-Up-Agent.

cs.AI↗

arXiv2025

DetailFusion: A Dual-branch Framework with Detail Enhancement for Composed Image Retrieval

Composed Image Retrieval (CIR) aims to retrieve target images from a gallery based on a reference image and modification text as a combined query. Recent approaches focus on balancing global information from two modalities and encode the query into a unified feature for retrieval. However, due to insufficient attention to fine-grained details, these coarse fusion methods often struggle with handling subtle visual alterations or intricate textual instructions. In this work, we propose DetailFusion, a novel dual-branch framework that effectively coordinates information across global and detailed granularities, thereby enabling detail-enhanced CIR. Our approach leverages atomic detail variation priors derived from an image editing dataset, supplemented by a detail-oriented optimization strategy to develop a Detail-oriented Inference Branch. Furthermore, we design an Adaptive Feature Compositor that dynamically fuses global and detailed features based on fine-grained information of each unique multimodal query. Extensive experiments and ablation analyses not only demonstrate that our method achieves state-of-the-art performance on both CIRR and FashionIQ datasets but also validate the effectiveness and cross-domain adaptability of detail enhancement for CIR.

cs.CV↗

arXiv2025

NTIRE 2025 challenge on Text to Image Generation Model Quality Assessment

This paper reports on the NTIRE 2025 challenge on Text to Image (T2I) generation model quality assessment, which will be held in conjunction with the New Trends in Image Restoration and Enhancement Workshop (NTIRE) at CVPR 2025. The aim of this challenge is to address the fine-grained quality assessment of text-to-image generation models. This challenge evaluates text-to-image models from two aspects: image-text alignment and image structural distortion detection, and is divided into the alignment track and the structural track. The alignment track uses the EvalMuse-40K, which contains around 40K AI-Generated Images (AIGIs) generated by 20 popular generative models. The alignment track has a total of 371 registered participants. A total of 1,883 submissions are received in the development phase, and 507 submissions are received in the test phase. Finally, 12 participating teams submitted their models and fact sheets. The structure track uses the EvalMuse-Structure, which contains 10,000 AI-Generated Images (AIGIs) with corresponding structural distortion mask. A total of 211 participants have registered in the structure track. A total of 1155 submissions are received in the development phase, and 487 submissions are received in the test phase. Finally, 8 participating teams submitted their models and fact sheets. Almost all methods have achieved better results than baseline methods, and the winning methods in both tracks have demonstrated superior prediction performance on T2I model quality assessment.

cs.CV↗

arXiv2025

Procedure of tuning up a three-site artificial Kitaev chain based on transmon measurements

Artificial Kitaev chains (AKCs), formed of quantum dot-superconductor linear arrays, provide a promising platform for hosting Majorana bound states (MBSs) and implementing topological quantum computing. The main challenges along this research direction would include the tuning up of AKCs for hosting MBSs and the readout of the parity of the chains. In this work, we present a step-by-step procedure for tuning up a three-site AKC to its sweet spots based on the spectra of a transmon circuit which is integrated with the chain for the purpose of reading out the parity of the chain. The signatures of the transmon's plasma modes in each step, particular those related to the appearance of MBSs in the chain, will be given. We find that the sweet spots in a three-site AKC can be classified into three types based on the relative strengths of elastic cotunneling (ECT) and crossed Andreev reflection (CAR): ECT-dominated sweet spots, genuine sweet spots and CAR-dominated sweet spots. We show that the ECT-dominated and CAR-dominated sweet spots can be more conveniently accessed and utilized in transmon-based measurements.

cond-mat.mes-hall↗