Arrow Research search

Author name cluster

Xiang-Yang Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

12 papers
1 author row

Possible papers

12

AAAI Conference 2025 Conference Paper

A-VL: Adaptive Attention for Large Vision-Language Models

  • Junyang Zhang
  • Mu Yuan
  • Ruiguang Zhong
  • Puhan Luo
  • Huiyou Zhan
  • Ningkang Zhang
  • Chengchen Hu
  • Xiang-Yang Li

The Large Vision-Language Model (LVLM) integrates computer vision and natural language processing techniques, offering substantial application potential. However, these models demand extensive resources during inference. Adaptive attention techniques can dynamically reduce computational redundancy and thus improve efficiency. Although current adaptive attention methods significantly reduce the memory requirements of Transformer-based language models, they are not tailored for LVLMs. We observe that LVLMs generate responses from both remote image tokens and local text tokens, and different modalities have different attention patterns. This observation inspires us to manage the attention for each modality separately. Specifically, for visual input, we store the cache of potentially useful information but only compute the most critical parts. For language input, we care more about local information. Based on our observation and analysis of vision-language attention patterns, we develop A-VL, a plug-and-play adaptive attention tailored for LVLM inference. Extensive evaluations on three vision-language tasks and five datasets show the effectiveness of our designs. Our approach A-VL outperforms existing adaptive attention methods in reducing memory usage and computational load without compromising performance.

PDF Details DOI

AAAI Conference 2025 Conference Paper

FFCG: Effective and Fast Family Column Generation for Solving Large-Scale Linear Program

  • Yi-Xiang Hu
  • Feng Wu
  • Shaoang Li
  • Yifang Zhao
  • Xiang-Yang Li

Column Generation (CG) is an effective and iterative algorithm to solve large-scale linear programs (LP). During each CG iteration, new columns are added to improve the solution of the LP. Typically, CG greedily selects one column with the most negative reduced cost, which can be improved by adding more columns at once. However, selecting all columns with negative reduced costs would lead to the addition of redundant columns that do not improve the objective value. Therefore, selecting the appropriate columns to add is still an open problem and previous machine-learning-based approaches for CG only add a constant quantity of columns per iteration due to the state-space explosion problem. To address this, we propose Fast Family Column Generation (FFCG) — a novel reinforcement-learning-based CG that selects a variable number of columns as needed in an iteration. Specifically, we formulate the column selection problem in CG as an MDP and design a reward metric that balances both the convergence speed and the number of redundant columns. In our experiments, FFCG converges faster on the common benchmarks and reduces the number of CG iterations by 77.1% for Cutting Stock Problem (CSP) and 84.8% for Vehicle Routing Problem with Time Windows (VRPTW), and a 71.4% reduction in computing time for CSP and 84.0% for VRPTW on average compared to several state-of-the-art baselines.

PDF Details DOI

IJCAI Conference 2025 Conference Paper

Guiding Large Language Models in Modeling Optimization Problems via Question Partitioning

  • Xiaotian Pan
  • Junhao Fang
  • Feng Wu
  • Sijia Zhang
  • Yi-Xiang Hu
  • Shaoang Li
  • Xiang-Yang Li

Optimization problems are ubiquitous across various domains, such as resource scheduling, production planning, and sales management. Traditionally, they are modeled manually, leading to inefficiencies due to difficulties in communication and collaboration between modeling and domain experts. The emergence of Large Language Models (LLMs) has made automated modeling possible. However, real-world applications are often large-scale and have numerous variables and constraints, limiting the applicability of existing methods. To address this, we propose PaMOP, a novel modeling framework based on LLMs, to model optimization problems automatically, given only natural language descriptions. Specifically, we extract and partition the problems using a tree structure, guiding the LLMs to model each set of constraints with self-augmented prompts, thus reducing the demands on the LLM's capabilities of large contents. The mathematical model is then iteratively corrected and validated through our correction procedures. The experiments demonstrate that our method improves performance on the common benchmark dataset NLP4LP, achieving an accuracy of 62. 3% and a code executability rate of 86. 8% when tested on GPT-4. Additionally, we demonstrate the effectiveness of our PaMOP in handling large real-world problems.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Model Lineage Closeness Analysis

  • Chen Tang
  • Lan Zhang
  • Qi Zhao
  • Xirong Zhuang
  • Xiang-Yang Li

As machine learning model modification techniques are extensively employed to obtain well-performing models at reduced costs, several studies have emerged to determine the presence of a modification relationship (i.e., lineage) between models. However, these methods are not robust to high-impact modification techniques and none of them have addressed the measurement of lineage closeness, which quantifies the degrees of modification. In this work, we visualize the changes in model decision boundaries resulting from different modification techniques and conclude that differences in decision boundaries serve as a precise metric of lineage closeness. Building upon this insight, we propose a modification-type agnostic and task-agnostic method to measure model lineage closeness by calculating mean adversarial distances from data points to decision boundaries and matching rate of data points, with data points selected through an efficient sampling method to reduce computational overhead. Moreover, we propose a novel indirect measurement approach to support lineage closeness measurement for models with different tasks. Finally, comprehensive experiments show that our design achieves an impressive 97% accuracy in lineage determination, and can precisely measure model lineage closeness for different modifications.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

Bandits with Concave Aggregated Reward

  • Yingqi Yu
  • Sijia Zhang
  • Shaoang Li
  • Lan Zhang
  • Wei Xie
  • Xiang-Yang Li

Multi-armed bandit is a simple but powerful algorithmic framework, and many effective algorithms have been proposed for various online models. In numerous applications, the decision-maker faces diminishing marginal utility. With non-linear aggregations, those algorithms often have poor regret bounds. Motivated by this, we study a bandit problem with diminishing marginal utility, which we termed the bandits with concave aggregated reward(BCAR). To tackle this problem, we propose two algorithms SW-BCAR and SWUCB-BCAR. Through theoretical analysis, we establish the effectiveness of these algorithms in addressing the BCAR issue. Extensive simulations demonstrate that our algorithms achieve better results than the most advanced bandit algorithms.

PDF Details DOI

AAAI Conference 2023 Conference Paper

SoftCorrect: Error Correction with Soft Detection for Automatic Speech Recognition

  • Yichong Leng
  • Xu Tan
  • Wenjie Liu
  • Kaitao Song
  • Rui Wang
  • Xiang-Yang Li
  • Tao Qin
  • Ed Lin

Error correction in automatic speech recognition (ASR) aims to correct those incorrect words in sentences generated by ASR models. Since recent ASR models usually have low word error rate (WER), to avoid affecting originally correct tokens, error correction models should only modify incorrect words, and therefore detecting incorrect words is important for error correction. Previous works on error correction either implicitly detect error words through target-source attention or CTC (connectionist temporal classification) loss, or explicitly locate specific deletion/substitution/insertion errors. However, implicit error detection does not provide clear signal about which tokens are incorrect and explicit error detection suffers from low detection accuracy. In this paper, we propose SoftCorrect with a soft error detection mechanism to avoid the limitations of both explicit and implicit error detection. Specifically, we first detect whether a token is correct or not through a probability produced by a dedicatedly designed language model, and then design a constrained CTC loss that only duplicates the detected incorrect tokens to let the decoder focus on the correction of error tokens. Compared with implicit error detection with CTC loss, SoftCorrect provides explicit signal about which words are incorrect and thus does not need to duplicate every token but only incorrect tokens; compared with explicit error detection, SoftCorrect does not detect specific deletion/substitution/insertion errors but just leaves it to CTC loss. Experiments on AISHELL-1 and Aidatatang datasets show that SoftCorrect achieves 26.1% and 9.4% CER reduction respectively, outperforming previous works by a large margin, while still enjoying fast speed of parallel generation.

PDF Details DOI

AAAI Conference 2022 Conference Paper

MLink: Linking Black-Box Models for Collaborative Multi-Model Inference

  • Mu Yuan
  • Lan Zhang
  • Xiang-Yang Li

The cost efficiency of model inference is critical to realworld machine learning (ML) applications, especially for delay-sensitive tasks and resource-limited devices. A typical dilemma is: in order to provide complex intelligent services (e. g. smart city), we need inference results of multiple ML models, but the cost budget (e. g. GPU memory) is not enough to run all of them. In this work, we study underlying relationships among black-box ML models and propose a novel learning task: model linking. Model linking aims to bridge the knowledge of different black-box models by learning mappings (dubbed model links) between their output spaces. Based on model links, we developed a scheduling algorithm, named MLink. Through collaborative multi-model inference enabled by model links, MLink can improve the accuracy of obtained inference results under the cost budget. We evaluated MLink on a multi-modal dataset with seven different ML models and two real-world video analytics systems with six ML models and 3, 264 hours of video. Experimental results show that our proposed model links can be effectively built among various black-box models. Under the budget of GPU memory, MLink can save 66. 7% inference computations while preserving 94% inference accuracy, which outperforms multi-task learning, deep reinforcement learning-based scheduler and frame filtering baselines.

PDF Details

AAAI Conference 2021 Conference Paper

Learning to Reweight with Deep Interactions

  • Yang Fan
  • Yingce Xia
  • Lijun Wu
  • Shufang Xie
  • Weiqing Liu
  • Jiang Bian
  • Tao Qin
  • Xiang-Yang Li

Recently, the concept of teaching has been introduced into machine learning, in which a teacher model is used to guide the training of a student model (which will be used in real tasks) through data selection, loss function design, etc. Learning to reweight, which is a specific kind of teaching that reweights training data using a teacher model, receives much attention due to its simplicity and effectiveness. In existing learning to reweight works, the teacher model only utilizes shallow/surface information such as training iteration number and loss/accuracy of the student model from training/validation sets, but ignores the internal states of the student model, which limits the potential of learning to reweight. In this work, we propose an improved data reweighting algorithm, in which the student model provides its internal states to the teacher model, and the teacher model returns adaptive weights of training samples to enhance the training of the student model. The teacher model is jointly trained with the student model using meta gradients propagated from a validation set. Experiments on image classification with clean/noisy labels and neural machine translation empirically demonstrate that our algorithm makes significant improvement over previous methods.

PDF Details

TCS Journal 2010 Journal Article

Mechanism design for set cover games with selfish element agents

  • Xiang-Yang Li
  • Zheng Sun
  • Weizhao Wang
  • Xiaowen Chu
  • Shaojie Tang
  • Ping Xu

In this article, we study the set cover games when the elements are selfish agents, each of which has a privately known valuation of receiving the service from the sets, i. e. , being covered by some set. Each set is assumed to have a fixed cost. We develop several approximately efficient strategyproof mechanisms that decide, after soliciting the declared bids by all elements, which elements will be covered, which sets will provide the coverage to these selected elements, and how much each element will be charged. For single-cover set cover games, we present a mechanism that is at least 1 d max -efficient, i. e. , the total valuation of all selected elements is at least 1 d max fraction of the total valuation produced by any mechanism. Here, d max is the maximum size of the sets. For multi-cover set cover games, we present a budget-balanced strategyproof mechanism that is 1 d max H d max -efficient under reasonable assumptions. Here, H n is the harmonic function. For the set cover games when both sets and elements are selfish agents, we show that a cross-monotonic payment-sharing scheme does not necessarily induce a strategyproof mechanism.

Details DOI

TCS Journal 2007 Journal Article

Average case analysis for tree labelling schemes

  • Ming-Yang Kao
  • Xiang-Yang Li
  • Weizhao Wang

We study how to label the vertices of a tree in such a way that we can decide the distance of two vertices in the tree given only their labels. Gavoille et al. proved that for any such distance labelling scheme, the maximum label length is at least 1 8 log 2 n − O ( log n ) bits, where n is the number of vertices in the input tree T. They also gave a separator-based labelling scheme that has the optimal label length Θ ( log n ⋅ log ( H n ( T ) ) ), where H n ( T ) is the height of T. We present two distance labelling schemes, namely, the backbone-based scheme and rake-based scheme, which also achieve the optimal label length. The two schemes always perform at least as well as the separator scheme. Furthermore, the rake-based scheme has a much smaller expected label length under certain tree distributions. With these new schemes, we also can find the least common ancestor of any two vertices based on their labels only.

Details DOI

TCS Journal 2003 Journal Article

Generating well-shaped d-dimensional Delaunay Meshes

  • Xiang-Yang Li

A d-dimensional simplicial mesh is a Delaunay triangulation if the circumsphere of each of its simplices does not contain any vertices inside. A mesh is well shaped if the maximum aspect ratio of all its simplices is bounded from above by a constant. It is a long-term open problem to generate well-shaped d-dimensional Delaunay meshes for a given polyhedral domain. In this paper, we present a refinement-based method that generates well-shaped d-dimensional Delaunay meshes for any PLC domain with no small input angles. Furthermore, we show that the generated well-shaped mesh has O(n) d-simplices, where n is the smallest number of d-simplices of any almost-good meshes for the same domain. Here a mesh is almost-good if each of its simplices has a bounded circumradius to the shortest edge length ratio.

Details DOI
v2026.09.13