Author name cluster

Yu Feng

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.

10 papers

2 author rows

AAAI Conference 2026 Conference Paper

FastAnimate: Towards Learnable Template Construction and Pose Deformation for Fast 3D Human Avatar Animation

Jian Shu
Nanjie Yao
Gangjian Zhang
Junlong Ren
Yu Feng
Hao Wang

3D human avatar animation aims at transforming a human avatar from an arbitrary initial pose to a specified target pose using deformation algorithms. Existing approaches typically divide this task into two stages: canonical template construction and target pose deformation. However, current template construction methods demand extensive skeletal rigging and often produce artifacts in contact regions. Moreover, target pose deformation suffers from structural distortions caused by Linear Blend Skinning (LBS), which significantly undermines animation realism. To address these problems, we propose a unified learning-based framework to address both challenges in two phases. For the former phase, to overcome the inefficiencies and artifacts during template construction, we leverage a U-Net architecture that decouples texture and pose information in a feed-forward process, enabling fast generation of a human template. For the latter phase, we propose a data-driven refinement technique that enhances structural integrity. Extensive experiments show that our model delivers consistent performance across diverse poses with an optimal balance between efficiency and quality, surpassing state-of-the-art (SOTA) methods.

PDF Details DOI

JBHI Journal 2026 Journal Article

Refocal Loss in Transformer for Long-Tailed Multi-Granularity Cataract Classification

Qiong Wang
Yan Wang
Hongdi Sun
Yu Feng
Zhe Dong
Cong Bai

Different cataract types and various severities usually require different countermeasures. For automatic cataract diagnosis, existing cataract classification methods group cataracts into common types, such as nuclear cataract, cortical cataract, and posterior subcapsular cataract, while existing cataract grading works aim to achieve fine-grained evaluation of the severity of the most common types of cataract. The severity assessment differs among various types of cataracts. Existing work is limited in predicting various cataract types at different granularity levels. In order to improve diagnostic efficiency, our study explores this matter in the context of multi-granularity cataract classification. Firstly, a large-scale dataset called Multi-Granularity Long-Tailed Cataract is collected. Secondly, an end-to-end training network is proposed, in which the Transformer is investigated for the extraction of multi-granularity cataract features. What is more, considering the imbalanced cataract data with the long-tailed distribution, the Refocal loss is proposed to rebalance the loss contribution of different classes by enhancing the reciprocal value of the effective number of samples. Compared with state-of-the-art methods, the experiments conducted on the multi-granularity cataract classification dataset demonstrate that the proposed model achieves the highest Precision of 78. 22%, F1-score of 68. 35%, Kappa of 64. 38% and MCC of 64. 49%, indicating that the proposed framework is promising in offering physicians reliable quantitative evaluations for multi-granularity cataract classification, which can help guide appropriate treatment decisions before the patient’s cataracts worsen.

Details DOI

NeurIPS Conference 2025 Conference Paper

ClusterFusion: Expanding Operator Fusion Scope for LLM Inference via Cluster-Level Collective Primitive

Xinhao Luo
Zihan Liu
Yangjie Zhou
Shihan Fang
Ziyu Huang
Yu Feng
Chen Zhang
Shixuan Sun

Large language model (LLM) decoding suffers from high latency due to fragmented execution across operators and heavy reliance on off-chip memory for data exchange and reduction. This execution model limits opportunities for fusion and incurs significant memory traffic and kernel launch overhead. While modern architectures such as NVIDIA Hopper provide distributed shared memory and low-latency intra-cluster interconnects, they expose only low-level data movement instructions, lacking structured abstractions for collective on-chip communication. To bridge this software-hardware gap, we introduce two cluster-level communication primitives, ClusterReduce and ClusterGather, which abstract common communication patterns and enable structured, high-speed data exchange and reduction between thread blocks within a cluster, allowing intermediate results to be on-chip without involving off-chip memory. Building on these abstractions, we design ClusterFusion, an execution framework that schedules communication and computation jointly to expand operator fusion scope by composing decoding stages such as QKV Projection, Attention, and Output Projection into a single fused kernels. Evaluations on H100 GPUs show that ClusterFusion outperforms state-of-the-art inference frameworks by $1. 61\times$ on average in end-to-end latency across different models and configurations.

PDF Details

ECAI Conference 2025 Conference Paper

Enhancing Few Shot Named Entity Recognition via Label Semantic Description and Diversity Text

Hui Wang 0170
Fang Du
Jiakun Li
Yu Feng

Large Language Models (LLMs) have demonstrated remarkable few-shot learning capabilities on Named Entity Recognition (NER) tasks, particularly through prompt-based approaches that avoid additional fine-tuning. However, despite the powerful generative and reasoning abilities of LLMs, two critical challenges remain: (1) semantic discrepancy of the same label across different datasets, which leads to recognition errors when using general labels to guide model outputs, and (2) contextual homogeneity in in-context examples, which limits the model’s ability to distinguish fine-grained entity types during inference. To address these challenges, we propose LSDNER, a dual-faceted prompt construction strategy that integrates structured label semantic descriptions and promotes context diversity. Specifically, to tackle label semantic inconsistency, we introduce a structured framework that organizes label semantic descriptions into definitions, attributes, relational features, and behavioral characteristics. This representation enables LLMs to better understand the dataset-specific semantic meanings of entity labels. To address contextual monotony, we devise a diversity-driven sampling strategy for selecting in-context demonstrations, thereby expanding semantic coverage and promoting reasoning capabilities. Experiments on three general and four domain-specific NER datasets demonstrate that our approach surpasses prompt-based methods and achieves competitive results with supervised baselines. Our code is available at: https: //github. com/hui68633/LSDNER.

Details

NeurIPS Conference 2025 Conference Paper

HyperGraphRAG: Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation

Haoran Luo
Haihong E
Guanting Chen
Yandan Zheng
Xiaobao Wu
Yikai Guo
Qika Lin
Yu Feng

Standard Retrieval-Augmented Generation (RAG) relies on chunk-based retrieval, whereas GraphRAG advances this approach by graph-based knowledge representation. However, existing graph-based RAG approaches are constrained by binary relations, as each edge in an ordinary graph connects only two entities, limiting their ability to represent the n-ary relations (n >= 2) in real-world knowledge. In this work, we propose HyperGraphRAG, the first hypergraph-based RAG method that represents n-ary relational facts via hyperedges. HyperGraphRAG consists of a comprehensive pipeline, including knowledge hypergraph construction, retrieval, and generation. Experiments across medicine, agriculture, computer science, and law demonstrate that HyperGraphRAG outperforms both standard RAG and previous graph-based RAG methods in answer accuracy, retrieval efficiency, and generation quality.

PDF Details

AAAI Conference 2025 Conference Paper

Incremental Nyström-based Multiple Kernel Clustering

Yu Feng
Weixuan Liang
Xinhang Wan
Jiyuan Liu
Suyuan Liu
Qian Qu
Renxiang Guan
Huiying Xu

Existing Multiple Kernel Clustering (MKC) algorithms commonly utilize the Nyström method to handle large-scale datasets. However, most of them employ uniform sampling for kernel matrix approximation, hence failing to accurately capture the underlying data structure, leading to large approximation errors. Additionally, they often use the same landmark points for all kernel matrix approximations, reducing kernel diversity. Moreover, in scenarios where approximate kernel matrices emerge over time, these methods require storing historical kernel information and recalculating, resulting in inefficient resource utilization. To address these issues, we propose a novel MKC algorithm, termed Incremental Nyström-based Multiple Kernel Clustering (INMKC). Specifically, leverage score sampling is utilized to reduce kernel approximation errors and enhance kernel diversity. Furthermore, we employ a consensus clustering structure that aligns with the newly emerged base kernel matrix for updates, avoiding recalculating previous kernel matrices, thus saving substantial computational resources. Additionally, we tackle the challenge of aligning incremental approximate kernels with different landmark points. Extensive experiments on the proposed INMKC demonstrate its effectiveness and efficiency compared to state-of-the-art methods.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Structure-Adaptive Multi-View Graph Clustering for Remote Sensing Data

Renxiang Guan
Wenxuan Tu
Siwei Wang
Jiyuan Liu
Dayu Hu
Chang Tang
Yu Feng
Junhong Li

Multi-view clustering (MVC) for remote sensing data is a critical and challenging task in Earth observation. Although recent advances in graph neural network (GNN)-based MVC have shown remarkable success, the most prevalent approaches have two major limitations: 1) heavily relying on a predefined yet fixed graph, which limits the performance of clustering because the large number of indistinguishable background samples contained in remote sensing data would introduce noise information and increase structure heterogeneity; 2) ignoring the effect of confusing samples on cluster structure compactness, which leads to fluffy cluster structure and decrease feature discriminability. To address these issues, we propose a Structure-Adaptive Multi-View Graph Clustering method named SAMVGC on remote sensing data which boosts the structure homogeneity and cluster compactness by adaptively learning the graph and cluster structures, respectively. Concretely, we use the geometric structure within the feature embedding space to refine adjacency matrices. The adjacency matrices are dynamically fused with the previous ones to improve the homogeneity and stability of structure information. Additionally, the samples are separated into two categories, including the central (intra-cluster center samples) and the confusing (inter-cluster boundary samples). On the basis, we deploy the contrastive learning paradigm on the central samples within views and the consistent learning paradigm on the confusing samples between views, improving the cluster compactness and consistency. Finally, we conduct extensive experiments on four benchmarks and achieve promising results, well demonstrating the effectiveness and superiority of the proposed method.

PDF Details DOI

EAAI Journal 2022 Journal Article

A Multi-Strategy Whale Optimization Algorithm and Its Application

Wenbiao Yang
Kewen Xia
Shurui Fan
Li Wang
Tiejun Li
Jiangnan Zhang
Yu Feng

Whale Optimization Algorithm (WOA) is a key tool for solving complex engineering optimization problems, aiming at adjusting important parameters to satisfy constraints and optimal objectives. WOA has a simple structure, few parameters, high search capability, and easy implementation. However, it suffers from the same problems as other metaheuristic algorithms of being prone to local optima and slow convergence, for which the Multi-Strategy Whale Optimization Algorithm (MSWOA) is proposed. Four strategies are introduced in MSWOA. Firstly, a highly randomized chaotic logistic map is used to generate a high-quality initial population. Secondly, exploitation and exploration are enhanced by setting adaptive weights and dynamic convergence factors. Further, a Lévy flight mechanism is introduced to maintain the population diversity in each iteration. Finally, the Evolutionary Population Dynamics (EPD) mechanism is introduced to improve the efficiency of search agents in finding the optimum. Another problem lies in the Semi-Supervised Extreme Learning Machine (SSELM) based on manifold regularization is an effective classification and regression model, but the random generation of input weights and hidden layer thresholds and the grid selection of hyperparameters lead to unsatisfactory classification performance. To this end, we developed the MSWOA-SSELM model, optimally selected the parameters of SSLEM using MSWOA, and applied it to logging layer recognition, which effectively improved the accuracy of logging interpretation. By comparing the experiments with 14 swarm intelligence algorithms on 18 benchmark test functions, the CEC2017 benchmark suite, and an engineering application problem, the experimental results show that MSWOA is significantly superior and effective in solving global optimization problems. Finally, the proposed MSWOA-SSELM is applied in three wells and outperforms other classification models in terms of Accuracy (ACC), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE). It obtained the best results with 96. 2567% ACC, MAE of 0. 0749, and RMSE of 0. 3870.

Details DOI

NeurIPS Conference 2006 Conference Paper

PG-means: learning the number of clusters in data

Yu Feng
Greg Hamerly

We present a novel algorithm called PG-means which is able to learn the number of clusters in a classical Gaussian mixture model. Our method is robust and efficient; it uses statistical hypothesis tests on one-dimensional projections of the data and model to determine if the examples are well represented by the model. In so doing, we are applying a statistical test for the entire model at once, not just on a per-cluster basis. We show that our method works well in difficult cases such as non-Gaussian data, overlapping clusters, eccentric clusters, high dimension, and many true clusters. Further, our new method provides a much more stable estimate of the number of clusters than existing methods.

PDF Details

ICRA Conference 2005 Conference Paper

Deadlock Avoidance Petri Net Controller for Manufacturing Systems with Multiple Resource Service

Keyi Xing
Xiajie Jin
Yu Feng

Many important Petri net-based methods have been presented to prevent deadlocks in automated manufacturing system with concurrent sequential processes. This paper addresses deadlock problems in automated manufacturing systems with multiple resource service. Petri net is used to model the flow of the parts and the usage and release of the resources. Deadlock structure objects in Petri net models, which lead the system to deadlock, are characterized. This paper proves that, by adding a control place for each deadlock structure so that the number of parts in the deadlock structure is limited, deadlock can be successfully prevented. Using the including relation between deadlock structures, from inside to outside, the Petri net controller is synthesized, and it is proved that the presented controller is a maximally primitive Petri net controller.

Details