Author name cluster

Xiangde Min

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.

5 papers

2 author rows

ICML Conference 2025 Conference Paper

FuseUNet: A Multi-Scale Feature Fusion Method for U-like Networks

Quansong He
Xiangde Min
Kaishen Wang
Tao He 0016

Medical image segmentation is a critical task in computer vision, with UNet serving as a milestone architecture. The typical component of UNet family is the skip connection, however, their skip connections face two significant limitations: (1) they lack effective interaction between features at different scales, and (2) they rely on simple concatenation or addition operations, which constrain efficient information integration. While recent improvements to UNet have focused on enhancing encoder and decoder capabilities, these limitations remain overlooked. To overcome these challenges, we propose a novel multi-scale feature fusion method that reimagines the UNet decoding process as solving an initial value problem (IVP), treating skip connections as discrete nodes. By leveraging principles from the linear multistep method, we propose an adaptive ordinary differential equation method to enable effective multi-scale feature fusion. Our approach is independent of the encoder and decoder architectures, making it adaptable to various U-Net-like networks. Experiments on ACDC, KiTS2023, MSD brain tumor, and ISIC2017/2018 skin lesion segmentation datasets demonstrate improved feature utilization, reduced network parameters, and maintained high performance. The code is available at https: //github. com/nayutayuki/FuseUNet.

Details

NeurIPS Conference 2025 Conference Paper

MobileODE: An Extra Lightweight Network

Le Yu
Jun Wu
Bo Gou
Xiangde Min
Lei Zhang
Zhang Yi
Tao He

Depthwise-separable convolution has emerged as a significant milestone in the lightweight development of Convolutional Neural Networks (CNNs) over the past decade. This technique consists of two key components: depthwise convolution, which captures spatial information, and pointwise convolution, which enhances channel interactions. In this paper, we propose a novel method to lightweight CNNs through the discretization of Ordinary Differential Equations (ODEs). Specifically, we optimize depthwise-separable convolution by replacing the pointwise convolution with a discrete ODE module, termed the \emph{\textbf{C}hannelwise \textbf{O}DE \textbf{S}olver (COS)}. The COS module is constructed by a simple yet efficient direct differentiation Euler algorithm, using learnable increment parameters. This replacement reduces parameters by over $98. 36$\% compared to conventional pointwise convolution. By integrating COS into MobileNet, we develop a new extra lightweight network called MobileODE. With carefully designed basic and inverse residual blocks, the resulting MobileODEV1 and MobileODEV2 reduce channel interaction parameters by $71. 0$\% and $69. 2$\%, respectively, compared to MobileNetV1, while achieving higher accuracy across various tasks, including image classification, object detection, and semantic segmentation. The code is available at {\url{https: //github. com/cashily/MobileODE}}.

PDF Details

JBHI Journal 2022 Journal Article

A Graph Convolutional Multiple Instance Learning on a Hypersphere Manifold Approach for Diagnosing Chronic Obstructive Pulmonary Disease in CT Images

Ling Chen
Qixing Feng
Xi Yin
Xiangde Min
Lei Shi
Defu Yang
Yen-Wei Chen
Daoqiang Zhang

Chronic obstructive pulmonary disease (COPD) is a prevalent chronic disease with high morbidity and mortality. The early diagnosis of COPD is vital for clinical treatment, which helps patients to have a better quality of life. Because COPD can be ascribed to chronic bronchitis and emphysema, lesions in a computed tomography (CT) image can present anywhere inside the lung with different types, shapes and sizes. Multiple instance learning (MIL) is an effective tool for solving COPD discrimination. In this study, a novel graph convolutional MIL with the adaptive additive margin loss (GCMIL-AAMS) approach is proposed to diagnose COPD by CT. Specifically, for those early stage patients, the selected instance-level features can be more discriminative if they were learned by our proposed graph convolution and pooling with self-attention mechanism. The AAMS loss can utilize the information of COPD severity on a hypersphere manifold by adaptively setting the angular margins to improve the performance, as the severity can be quantified as four grades by pulmonary function test. The results show that our proposed GCMIL-AAMS method provides superior discrimination and generalization abilities in COPD discrimination, with areas under a receiver operating characteristic curve (AUCs) of 0. 960 $\pm$ 0. 014 and 0. 862 $\pm$ 0. 010 in the test set and external testing set, respectively, in 5-fold stratified cross validation; moreover, it demonstrates that graph learning is applicable to MIL and suggests that MIL may be adaptable to graph learning.

Details DOI

JBHI Journal 2022 Journal Article

Brain Fingerprinting and Lie Detection: A Study of Dynamic Functional Connectivity Patterns of Deception Using EEG Phase Synchrony Analysis

Junfeng Gao
Lingyun Gu
Xiangde Min
Pan Lin
Chenhong Li
Quan Zhang
Nini Rao

This study investigated the brain functional connectivity (FC) patterns related to lie detection (LD) tasks with the purpose of analyzing the underlying cognitive processes and mechanisms in deception. Using the guilty knowledge test protocol, 30 subjects were divided randomly into guilty and innocent groups, and their electroencephalogram (EEG) signals were recorded on 32 electrodes. Phase synchrony of EEG was analyzed between different brain regions. A few-trials-based relative phase synchrony (FTRPS) measure was proposed to avoid the false synchronization that occurs due to volume conduction. FTRPS values with a significantly statistical difference between two groups were employed to construct FC patterns of deception, and the FTRPS values from the FC networks were extracted as the features for the training and testing of the support vector machine. Finally, four more intuitive brain fingerprinting graphs (BFG) on delta, theta, alpha and beta bands were respectively proposed. The experimental results reveal that deceptive responses elicited greater oscillatory synchronization than truthful responses between different brain regions, which plays an important role in executing lying tasks. The functional connectivity in the BFG is mainly implicated in the visuo-spatial imagery, bottom-top attention and memory systems, work memory and episodic encoding, and top-down attention and inhibition processing. These may, in part, underlie the mechanism of communication between different brain cortices during lying. High classification accuracy demonstrates the validation of BFG to identify deception behavior, and suggests that the proposed FTRPS could be a sensitive measure for LD in the real application.

Details DOI

JBHI Journal 2022 Journal Article

Effective Connectivity in Cortical Networks During Deception: A Lie Detection Study Based on EEG

Junfeng Gao
Xiangde Min
Qianruo Kang
Huifang Si
Huimiao Zhan
Anne Manyande
Xuebi Tian
Yinhong Dong

Thus far, when deception behaviors occur, the connectivity patterns and the communication between different brain areas remain largely unclear. In this study, the most important information flows (MIIFs) between different brain cortices during deception were explored. First, the guilty knowledge test protocol was employed, and 64 electrodes’ electroencephalogram (EEG) signals were recorded from 30 subjects (15 guilty and 15 innocent). Cortical current density waveforms were then estimated on the 24 regions of interest (ROIs). Next, partial directed coherence (PDC), an effective connectivity (EC) analysis was applied in the cortical waveforms to obtain the brain EC networks for four bands: delta (1–4 Hz), theta (4–8 Hz), alpha (8–13 Hz) and beta (13–30 Hz). Furthermore, using the graph theoretical analysis, the network parameters with significant differences in the EC network were extracted as features to identify the two groups. The high classification accuracy of the four bands demonstrated that the proposed method was suitable for lie detection. In addition, based on the optimal features in the classification mode, the brain “hub” regions were identified, and the MIIFs were significantly different between the guilty and innocent groups. Moreover, the fronto-parietal network was found to be most prominent among all MIIFs at the four bands. Furthermore, combining the neurophysiology significance of the four frequency bands, the roles of all MIIFs were analyzed, which could help us to uncover the underlying cognitive processes and mechanisms of deception.

Details DOI