Author name cluster

Lingfeng Wang

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.

7 papers

1 author row

NeurIPS Conference 2025 Conference Paper

ALTo: Adaptive-Length Tokenizer for Autoregressive Mask Generation

Lingfeng Wang
Hualing Lin
Senda Chen
Tao Wang
Changxu Cheng
Yangyang Zhong
Dong Zheng
Wuyue Zhao

While humans effortlessly draw visual objects and shapes by adaptively allocating attention based on their complexity, existing multimodal large language models (MLLMs) remain constrained by rigid token representations. Bridging this gap, we propose ALTo, an adaptive length tokenizer for autoregressive mask generation. To achieve this, a novel token length predictor is designed, along with a length regularization term and a differentiable token chunking strategy. We further build ALToLLM that seamlessly integrates ALTo into MLLM. Preferences on the trade-offs between mask quality and efficiency is implemented by group relative policy optimization (GRPO). Experiments demonstrate that ALToLLM achieves state-of-the-art performance with adaptive token cost on popular segmentation benchmarks. Code and models will be released.

PDF Details

AAAI Conference 2025 Conference Paper

Progressive Self-Learning for Domain Adaptation on Symbolic Regression of Integer Sequences

Yaohui Zhu
Kaiming Sun
Zhengdong Luo
Lingfeng Wang

Symbolic Regression of Integer Sequences (SRIS) aims to discover precise mathematical formulas from integer sequences. The neural machine translation-based method of SRIS trains the model using randomly generated data, and directly utilizes the trained model for inference on target sequences. However, the method often fails to effectively generalize to the target sequence, since the randomly generated data can not adequately cover the distributions of target data, i.e., there are distribution differences between them. In this work, we propose a progressive self-learning (PSL) method to explicitly capture sequence-formula distributions of the target domain. Specifically, a source domain dataset is generated by incorporating initial terms of the target domain to reduce the sequence distribution gap between the source domain and the target domain. Meanwhile, a self-learning loop strategy is adopted to improve the ability of the model to capture the sequence-formula distribution of the target domain. In this strategy, a neural machine translation model is used to learn the mappings from sequences to formulas in an end-to-end fashion. Then, this model is employed to explore candidate formulas of the target sequence using beam search. After verifying these candidate formula correctness, some of them are retained as training data for the next learning. Experimental results on OEIS datasets demonstrate that the proposed method surpasses current state-of-the-art methods in accuracy, and also discovers new formulas.

PDF Details DOI

AAAI Conference 2022 Conference Paper

Learning from the Target: Dual Prototype Network for Few Shot Semantic Segmentation

Binjie Mao
Xinbang Zhang
Lingfeng Wang
Qian Zhang
Shiming Xiang
Chunhong Pan

Due to the scarcity of annotated samples, the diversity between support set and query set becomes the main obstacle for few shot semantic segmentation. Most existing prototypebased approaches only exploit the prototype from the support feature and ignore the information from the query sample, failing to remove this obstacle. In this paper, we proposes a dual prototype network (DPNet) to dispose of few shot semantic segmentation from a new perspective. Along with the prototype extracted from the support set, we propose to build the pseudo-prototype based on foreground features in the query image. To achieve this goal, the cycle comparison module is developed to select reliable foreground features and generate the pseudo-prototype with them. Then, a prototype interaction module is utilized to integrate the information of the prototype and the pseudo-prototype based on their underlying correlation. Finally, a multi-scale fusion module is introduced to capture contextual information during the dense comparison between prototype (pseudo-prototype) and query feature. Extensive experiments conducted on two benchmarks demonstrate that our method exceeds previous state-of-thearts with a sizable margin, verifying the effectiveness of the proposed method.

PDF Details

AAAI Conference 2019 Conference Paper

What and Where the Themes Dominate in Image

Xinyu Xiao
Lingfeng Wang
Shiming Xiang
Chunhong Pan

The image captioning is to describe an image with natural language as human, which has benefited from the advances in deep neural network and achieved substantial progress in performance. However, the perspective of human description to scene has not been fully considered in this task recently. Actually, the human description to scene is tightly related to the endogenous knowledge and the exogenous salient objects simultaneously, which implies that the content in the description is confined to the known salient objects. Inspired by this observation, this paper proposes a novel framework, which explicitly applies the known salient objects in image captioning. Under this framework, the known salient objects are served as the themes to guide the description generation. According to the property of the known salient object, a theme is composed of two components: its endogenous concept (what) and the exogenous spatial attention feature (where). Specifically, the prediction of each word is dominated by the concept and spatial attention feature of the corresponding theme in the process of caption prediction. Moreover, we introduce a novel learning method of Distinctive Learning (DL) to get more specificity of generated captions like human descriptions. It formulates two constraints in the theme learning process to encourage distinctiveness between different images. Particularly, reinforcement learning is introduced into the framework to address the exposure bias problem between the training and the testing modes. Extensive experiments on the COCO and Flickr30K datasets achieve superior results when compared with the state-of-the-art methods.

PDF Details

NeurIPS Conference 2018 Conference Paper

Structure-Aware Convolutional Neural Networks

Jianlong Chang
Jie Gu
Lingfeng Wang
Gaofeng Meng
Shiming Xiang
Chunhong Pan

Convolutional neural networks (CNNs) are inherently subject to invariable filters that can only aggregate local inputs with the same topological structures. It causes that CNNs are allowed to manage data with Euclidean or grid-like structures (e. g. , images), not ones with non-Euclidean or graph structures (e. g. , traffic networks). To broaden the reach of CNNs, we develop structure-aware convolution to eliminate the invariance, yielding a unified mechanism of dealing with both Euclidean and non-Euclidean structured data. Technically, filters in the structure-aware convolution are generalized to univariate functions, which are capable of aggregating local inputs with diverse topological structures. Since infinite parameters are required to determine a univariate function, we parameterize these filters with numbered learnable parameters in the context of the function approximation theory. By replacing the classical convolution in CNNs with the structure-aware convolution, Structure-Aware Convolutional Neural Networks (SACNNs) are readily established. Extensive experiments on eleven datasets strongly evidence that SACNNs outperform current models on various machine learning tasks, including image classification and clustering, text categorization, skeleton-based action recognition, molecular activity detection, and taxi flow prediction.

PDF Details

EAAI Journal 2013 Journal Article

Intelligent state space pruning for Monte Carlo simulation with applications in composite power system reliability

Robert C. Green
Lingfeng Wang
Mansoor Alam
Chanan Singh

The probabilistic reliability evaluation of composite power systems is a complicated, computation intensive, and combinatorial task. As such evaluation may suffer from issues regarding high dimensionality that lead to an increased need for computational resources, MCS is often used to evaluate the reliability of power systems. In order to alleviate this burden, an analytical method known as state space decomposition has previously been used to prune the state space that is sampled using MCS. This paper extends the state-of-the-art by proposing a novel algorithm known as intelligent state space pruning (ISSP). This algorithm leverages the intelligence of highly modified population based metaheuristic (PBM) algorithms including genetic algorithms (GA), particle swarm optimization (PSO), ant colony optimization (ACO), and artificial immune systems (AIS) to quickly, efficiently, and intelligently prune the state space that is used during MCS. The presented PBMs are modified using domain-specific knowledge to improve their performance and fine tune their intelligence. This new algorithm leads to reductions of up to 90% in total computation time and iterations required for convergence when compared to non-sequential MCS. Results are reported using the IEEE Reliability Test Systems (RTS79/MRTS).

Details DOI

EAAI Journal 2009 Journal Article

Reserve-constrained multiarea environmental/economic dispatch based on particle swarm optimization with local search

Lingfeng Wang
Chanan Singh

The objective of economic dispatch (ED) is to minimize the total operational cost while satisfying the operational constraints of power systems. Multiarea economic dispatch (MAED) deals with the optimal power dispatch of multiple areas. In this investigation, multiarea environmental/economic dispatch (MAEED) is proposed to address the environmental issue during the ED. Its target is to dispatch the power among different areas by simultaneously minimizing the operational costs and pollutant emissions. In this paper, the MAEED problem is first formulated and then an improved multiobjective particle swarm optimization (MOPSO) algorithm is developed to derive a set of Pareto-optimal solutions. In the proposed version of MOPSO, local search is used to increase its search efficiency. To ensure the system security, tie-line transfer limits between different areas are incorporated as a set of constraints in the optimization process. Moreover, the reserve-sharing scheme is applied to ensure that each area is able to fulfill its reserve requirement. Numerical studies based on a four-area test power generation system are carried out to demonstrate the validity of the proposed optimization method as well as the results from different problem formulations. Comparative results with respect to other optimization methods are also presented.

Details DOI