Author name cluster

Jie Sun

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.

22 papers

2 author rows

EAAI Journal 2026 Journal Article

A lightweight cross-scale reconstruction framework with synergistic attention for insulator and defect detection in power grid environments

Entuo Li
Qinglong Wang
Rongwei Liu
Yongbao Chen
Jianghui Meng
Yunjian Hu
Wen Peng
Jie Sun

With the widespread adoption of Unmanned Aerial Vehicle (UAV) technology in power-grid inspection, automated and real-time detection of power insulators and their defects increasingly require lightweight and robust deep learning solutions. However, existing deep learning frameworks still face significant challenges in real inspection environments, where cluttered backgrounds, dense object distributions, frequent occlusion, and substantial scale variation hinder both lightweight design and robust detection performance. To address these challenges, we propose a lightweight single-stage detection framework for insulators and defects. The proposed framework combines cross-scale feature reconstruction with synergistic attention modeling to enhance semantic consistency across feature levels and improve the representation of small, dense, and occluded targets while preserving computational efficiency. Comprehensive evaluations demonstrate robust and competitive performance across multiple datasets. On the Complex Power-Grid Multi-Scenario Insulator Dataset (CPMID), which covers transmission lines, substations, and power plants, the proposed framework achieves a mean average precision (mAP) of 78. 93% at an Intersection over Union (IoU) threshold of 0. 5. It also achieves mAP values of 99. 34% and 99. 12% at the same IoU threshold on the Chinese Power Line Insulator Dataset (CPLID) and the Insulator Defect Detection Dataset (IDID), respectively, with only 3. 37 million parameters. This work provides an efficient and generalizable solution for automated insulator defect detection in complex power-grid environments.

Details DOI

EAAI Journal 2026 Journal Article

Adversarial attack-defense framework for enhancing the robustness of power insulator detection in cloud-edge deployment

Qinglong Wang
Changyu Yang
Jianhua Du
Na Li
Jian Wang
Huilong Han
Yunjian Hu
Wen Peng

Power insulators are essential to the stability of high-voltage transmission lines, and unmanned aerial vehicle-based inspection systems supported by cloud-edge computing have markedly improved monitoring efficiency. However, deep neural networks remain vulnerable to adversarial attacks. Most studies focus on image classification, with limited attention given to oriented bounding box (OBB) detection under edge constraints critical for power-line monitoring. To address this, we propose an adversarial attack-defense framework tailored for cloud-edge insulator inspection. We develop an enhanced Projected Gradient Descent Plus (PGD+) algorithm that integrates dynamic step-size scheduling, multi-point initialization, momentum accumulation, and stochastic gradient perturbation to generate effective yet visually imperceptible adversarial examples. Ablation analysis quantifies the contribution of each mechanism and demonstrates reliable transferability across different detectors. We construct a specialized glass insulator dataset with OBB annotations and design a dual-stage defense combining an attention-enhanced denoising autoencoder and an embedded denoising module integrated into the detector. Experiments show that PGD + achieves a 90. 6 % attack success rate, reducing mean average precision at an intersection-over-union threshold of 0. 5 (mAP50) by over 90 %, while preserving high visual fidelity with peak signal-to-noise ratios above 38 decibels (dB) and structural-similarity index values exceeding 0. 94. The joint defense recovers mAP50 to 93. 6 % under PGD + attacks. Edge-deployment evaluation on an embedded platform demonstrates that the full pipeline sustains 23–35 frames per second, demonstrating its practical feasibility for real-time cloud-edge inspection. The proposed framework delivers algorithmic strategies, a benchmark OBB dataset, and a deployable defense architecture that enhances the robustness of intelligent power-grid inspection in adversarial environments.

Details DOI

EAAI Journal 2026 Journal Article

Cross-scale recognition of dense insulators and defects in complex power grid environments

Qinglong Wang
Entuo Li
Shihao Cui
Wengang Yang
Xinchun Zhang
Wenqiang Jiang
Yunjian Hu
Wen Peng

As intelligent power grids continue to evolve, automated insulator inspection requires robust deep learning solutions capable of handling diverse operational scenarios. Current detection methods face substantial challenges within multi-scenario power grid environments, including varied insulator types, complex background interference, scale variation, and dense object overlap with occlusion. To address these limitations, we propose the Universal Power Insulator Detection Network (UPIDNet), an enhanced one-stage anchor-based architecture employing integrated cross-scale detection. The network incorporates a contextual transformer attention embedded within a cross-stage partial bottleneck structure with three convolutions to strengthen feature extraction. A four-scale feature pyramid enables integrated cross-scale fusion, mitigating background interference and accommodating multi-scale targets. The design features a high-resolution detection head optimized for small target detection and a channel and position attention module to expand receptive fields for improved spatial awareness. We further construct the Complex Power-grid Multi-scenario Insulator Dataset (CPMID), comprising 7712 images with 72, 596 annotated instances collected from transmission lines, substations, and power plants. Both the dataset and trained models are publicly available. Comprehensive evaluations demonstrate robust performance across multiple benchmarks. UPIDNet achieves a mean Average Precision (mAP) of 85. 94 % at an Intersection over Union (IoU) threshold of 0. 5 on CPMID, while maintaining competitive performance on two public insulator datasets. These contributions provide an effective and generalizable solution for automated insulator inspection in complex power grid environments.

Details DOI

AAAI Conference 2026 Conference Paper

Transferring Causal Driving Patterns for Generalizable Traffic Simulation with Diffusion-Based Distillation

Yuhang Chen
Jie Sun
Jialin Fan
Jian Sun

Traffic simulation is essential for validating the safety and reliability of autonomous driving systems, yet data-driven simulation methods often struggle with distribution shifts, limiting their generalizability across diverse datasets (domains). To address this, we present Causal Driving Pattern Transfer (CDPT), a novel two-stage knowledge distillation framework built upon diffusion model to enhance cross-domain generalizability. In Phase I, we implement hybrid self-distillation within the source domain by integrating feature-, response-, and contrastive-level distillation, which enables the model to decompose complex driving behaviors into their core causal components, including scene-conditioned driven patterns, multi-agent interaction dynamics and casual saliency. In Phase II, we introduce a continual distillation strategy: few-shot samples from the target domain are used to initiate generation of diverse synthetic scenarios, allowing the student model to continually adapt to novel environments without retraining on large-scale data. Extensive experiments demonstrate that CDPT achieves strong generalization in both open-loop and closed-loop simulations, effectively generating realistic, interaction-aware behaviors that are critical for scalable and reliable autonomous driving testing.

PDF Details DOI

ECAI Conference 2025 Conference Paper

A Self-Adaptive Frequency Domain Network for Continuous Intraoperative Hypotension Prediction

Xian Zeng
Tianze Xu
Kai Yang
Jie Sun
Youran Wang
Jun Xu
Mucheng Ren

Intraoperative hypotension (IOH) is strongly associated with postoperative complications, including postoperative delirium and increased mortality, making its early prediction crucial in perioperative care. While several artificial intelligence-based models have been developed to provide IOH warnings, existing methods face limitations in incorporating both time and frequency domain information, capturing short- and long-term dependencies, and handling noise sensitivity in biosignal data. To address these challenges, we propose a novel Self-Adaptive Frequency Domain Network (SAFDNet). Specifically, SAFDNet integrates an adaptive spectral block, which leverages Fourier analysis to extract frequency-domain features and employs self-adaptive thresholding to mitigate noise. Additionally, an interactive attention block is introduced to capture both long-term and short-term dependencies in the data. Extensive internal and external validations on two large-scale real-world datasets demonstrate that SAFDNet achieves up to 97. 3% AUROC in IOH early warning, outperforming state-of-the-art models. Furthermore, SAFDNet exhibits robust predictive performance and low sensitivity to noise, making it well-suited for practical clinical applications.

Details

AAAI Conference 2025 Conference Paper

IOP: An Idempotent-Like Optimization Method on the Pareto Front of Hypernetwork

Hui Wang
Renyu Yang
Jie Sun
Hao Peng
Xudong Mou
Tianyu Wo
Xudong Liu

Pareto Front Learning (PFL) has been one of the effective means to resolve multi-objective optimization problems through exploring all optimal solutions to learn the entire Pareto front. Pareto Hypernetwork (PHN) is a new promising way to generate the sequence of Pareto-optimal solutions that can be further used as potential solutions to constitute the Pareto front. However, the existing PHN-based approaches suffer from two performance issues: They take as inputs human-crafted preference vector or chunk embedding, rather than the input data samples, and thus vulnerable to data distribution shifts. Such approaches cannot optimize all potential solutions when forming the Pareto front, as they merely optimize the loss pertaining to one single input at a time of optimization round. To improve the quality of the Pareto front, we propose IOP, a novel Idempotent-like Optimization method to learn the entire Pareto front accurately and enhance Hypernetwork's adaptability to distribution shifts. In particular, IOP performs idempotent-like optimization by exploiting manifold space mapping, so that the target networks generated by the optimized Hypernetwork can effectively handle samples with similar distributions of the input samples, without the pre-defined human-crafted inputs. IOP maximizes the Hypervolume indicator that is composed of all potential solutions at a higher level. Experimental results demonstrate that IOP outperforms the state-of-the-art methods by 4.7% on average in producing the Pareto front and has a 10.5% improvement in adaptability.

PDF Details DOI

JBHI Journal 2025 Journal Article

Multi-Omics Deep-Learning Prediction of Homologous Recombination Deficiency-Like Phenotype Improved Risk Stratification and Guided Therapeutic Decisions in Gynecological Cancers

Yibo Zhang
Congcong Yan
Zijian Yang
Meng Zhou
Jie Sun

Homologous recombination deficiency (HRD) is a well-recognized important biomarker in determining the clinical benefits of platinum-based chemotherapy and PARP inhibitor therapy for patients diagnosed with gynecologic cancers. Accurate prediction of HRD phenotype remains challenging. Here, we proposed a novel Multi-Omics integrative Deep-learning framework named MODeepHRD for detecting HRD-positive phenotype. MODeepHRD utilizes a convolutional attention autoencoder that effectively leverages omics-specific and cross-omics complementary knowledge learning. We trained MODeepHRD on 351 ovarian cancer (OV) patients using transcriptomic, DNA methylation and mutation data, and validated it in 2133 OV samples of 22 datasets. The predicted HRD-positive tumors were significantly associated with improved survival (HR = 0. 68; 95% CI, 0. 60–0. 77; log-rank p < 0. 001 for meta-cohort; HR = 0. 5; 95% CI, 0. 29–0. 86; log-rank p = 0. 01 for ICGC-OV cohort) and higher response to platinum-based chemotherapy compared to predicted HRD-negative tumors. The translational potential of MODeepHRDs was further validated in multicenter breast and endometrial cancer cohorts. Furthermore, MODeepHRD outperforms conventional machine-learning methods and other similar task approaches. In conclusion, our study demonstrates the promising value of deep learning as a solution for HRD testing in the clinical setting. MODeepHRD holds potential clinical applicability in guiding patient risk stratification and therapeutic decisions, providing valuable insights for precision oncology and personalized treatment strategies.

Details DOI

EAAI Journal 2025 Journal Article

Signal-to-noise ratio guided noise adaptive network via Dual-domain collaboration for low-light image enhancement

Yifei Wang
Chunyan She
Jie Sun
Shukai Duan
Lidan Wang

Low-light image enhancement is crucial for accurate perception and decision-making, particularly in fields like autonomous driving. However, noise is inherently present in low-light images, especially in extremely dark regions, which complicates the reconstruction of clear images. Transformer, which typically computes self-attention scores across all available tokens, often struggle with the interference caused by this noise. To address this challenge, we propose the Signal-to-Noise Ratio (SNR) guided Noise Adaptive Network (SNA-Net), a novel approach that leverages the strengths of both Convolutional Neural Networks (CNN) and Transformer to adapt to noise distribution across different regions for low-light image enhancement. The SNA-Net introduces two key components within the transformer block: Noise Adaptive Self-Attention (NASA) and Dual-domain Refinement Feed-forward Network (DRFN) Specifically, NASA adaptively computes attention scores using both dense and sparse branches. The sparse branch filters out negative token interactions in low SNR regions, while the dense branch preserves essential image information. In parallel, DRFN reduces feature redundancy in both the spatial and frequency domains, thereby improving the recovery of the underlying clear image. Additionally, to facilitate better integration between CNN and Transformer features, we design an SNR-guided Feature Fusion Module (SGFF). We validate the superior performance of SNA-Net on six datasets through extensive experiments. Our code is available at https: //github. com/Wyyff993/SNA-NET.

Details DOI

JBHI Journal 2025 Journal Article

SkipDAEformer: A High-Precision Representation Learning Method for Removing Random Mixed Noise in MCG Signals

Ruizhe Wang
Zhanyi Liu
Jiaojiao Pang
Jie Sun
Min Xiang
Xiaolin Ning

Automated analytical techniques for magnetocardiography (MCG) are essential for diagnosing and predicting cardiovascular diseases. Clinically acquired MCG signals are often contaminated by various types of noise, which negatively impact subsequent signal analysis. However, traditional methods have limitations in denoising long-term MCG signals with complex spatial structures. We propose a high-precision, robust representation learning method based on skip connection multi-scale feature fusion (SkipDAEformer) for effectively removing random mixed noise in MCG signals. SkipDAEformer integrates attention fusion mechanisms into a basic denoising autoencoder to extract and fuse critical temporal and spatial information from each feature map, thus enhancing the model’s ability to capture long-range dependencies and spatial features in MCG signals. Meanwhile, we further supplement and refine the semantic information for the feature maps through a global feature fusion method. By fusing multi-scale features from different skip connections, SkipDAEformer can learn more comprehensive representations of MCG signals, enabling the effective separation of clean signals from noise. Experimental results demonstrate that SkipDAEformer outperforms existing methods in denoising performance, channel consistency, feature consistency, and generalization ability and can be extended to a self-supervised learning framework. In actual noise reduction and diagnostic classification tasks, SkipDAEformer shows superior clinical acceptability and diagnostic value, potentially advancing MCG data analysis.

Details DOI

EAAI Journal 2024 Journal Article

A novel deep ensemble reinforcement learning based control method for strip flatness in cold rolling steel industry

Wen Peng
Jiawei Lei
Chengyan Ding
Chongxiang Yue
Gengsheng Ma
Jie Sun
Dianhua Zhang

The flatness control system of cold-rolled strip is characterized by nonlinearity, strong coupling, and multivariable features. Theoretical models of flatness derived from physical mechanisms of the rolling process are intricate. This complexity makes it challenging to establish an accurate control model. Reinforcement learning (RL) is a machine learning method that uses an intelligent agent to learn from data interacting with its environment. In this study, considering the large quantity of historical data generated by the strip steel control system, we used the RL method to analyze the data offline in order to train a data-driven controller. We combined it with ensemble learning to propose a novel method for flatness control based on ensemble deep reinforcement learning (EDRL). Initially, we constructed a state space model for the cold-rolled flatness utilizing historical production data. This model served as the environment for offline RL training. We present an EDRL-based flatness control method that incorporates the principles of ensemble learning. Ensemble learning amalgamates multiple individual models to achieve superior performances. In the online simulation of rolling production, the EDRL-controlled flatness value was less than 5 (I-Unit) IU from the target flatness value, exceeding the performance of the proportional–integral (PI) controller. The proposed method surpassed the conventional approach in terms of both robustness and its capability to regulate flatness.

Details DOI

EAAI Journal 2024 Journal Article

An efficient detector for detecting surface defects on cold-rolled steel strips

Shuzong Chen
Shengquan Jiang
Xiaoyu Wang
Pu Sun
Changchun Hua
Jie Sun

Surface-defect inspection is vital in cold-rolled steel-strip manufacturing, given the complexities of production environments and the high speeds involved. Further, the defects on cold-rolled steel strips are often characterized by their small size, diversity of types, and similarities among different types, posing significant challenges in balancing detection accuracy and efficiency. To address the challenges, we designed a detector based on You Only Look Once version 5 (YOLOv5) to achieve precise detection of surface defects on cold-rolled steel strips. First, a dataset containing seven types of defects was curated, named the Cold-Rolled Steel Defect Dataset (CR7-DET). Next, a feature-extraction network based on residual-like connections within a single residual block (Res2net) was developed to enhance the model’s feature-extraction capability, alongside introducing a multi-head attention module to focus on key information features. To reduce the information loss during feature fusion, we established an adaptive feature-fusion Path Aggregation Network (aff-PAN), which was optimized by designing a lightweight adaptive down-sampling module (LAD) to increase the sensory-field implementation of feature fusion. The ghost convolution effectively reduced the number of parameters and increased the speed without affecting the model’s performance. Finally, experiments were conducted on our CR7-DET and a public dataset (GC10-DET). With a reduced parameter count of 6. 85 million, our model achieved a mean average precision(mAP) of 87. 6% on CR7-DET and 79. 7% on GC10-DET. The experimental results demonstrated that our model achieved a balance between detection accuracy and inference efficiency. The model has the potential to reduce scrap rates caused by defects and improve the overall surface quality of cold-rolled steel strips.

Details DOI

AAAI Conference 2024 Conference Paper

CREAD: A Classification-Restoration Framework with Error Adaptive Discretization for Watch Time Prediction in Video Recommender Systems

Jie Sun
Zhaoying Ding
Xiaoshuang Chen
Qi Chen
Yincheng Wang
Kaiqiao Zhan
Ben Wang

The watch time is a significant indicator of user satisfaction in video recommender systems. However, the prediction of watch time as a target variable is often hindered by its highly imbalanced distribution with a scarcity of observations for larger target values and over-populated samples for small values. State-of-the-art watch time prediction models discretize the continuous watch time into a set of buckets in order to consider the distribution of watch time. However, it is highly uninvestigated how these discrete buckets should be created from the continuous watch time distribution, and existing discretization approaches suffer from either a large learning error or a large restoration error. To address this challenge, we propose a Classification-Restoration framework with Error-Adaptive-Discretization (CREAD) to accurately predict the watch time. The proposed framework contains a discretization module, a classification module, and a restoration module. It predicts the watch time through multiple classification problems. The discretization process is a key contribution of the CREAD framework. We theoretically analyze the impacts of the discretization on the learning error and the restoration error, and then propose the error-adaptive discretization (EAD) technique to better balance the two errors, which achieves better performance over traditional discretization approaches. We conduct detailed offline evaluations on a public dataset and an industrial dataset, both showing performance gains through the proposed approach. Moreover, We have fully launched our framework to an online video platform, which resulted in a significant increase in users' video watch time by 0.29% through A/B testing. These results highlight the effectiveness of the CREAD framework in watch time prediction in video recommender systems.

PDF Details DOI

EAAI Journal 2024 Journal Article

Enhanced recognition of insulator defects on power transmission lines via proposal-based detection model with integrated improvement methods

Qinglong Wang
Yongjian Li
Shihao Cui
Na Li
Xinchun Zhang
Wenqiang Jiang
Wen Peng
Jie Sun

Deep learning-driven transmission line inspection is a critical area for smart power grid development. Despite advances in deep learning for insulator defect detection, challenges remain in model robustness and adaptability for the varying real-world adaptability, especially for insignificant defects in complex backgrounds. This study presents a comprehensive improvement strategy for detecting insulators and cross-scale broken defects on transmission lines, employing a proposal-based detection model. The model introduces a holistic pipeline of improved methods, including backbone modification, anchor box scale recalibration, and improvements in Region of Interest (RoI) downsampling alignment and Intersection over Union (IoU) loss function. Various backbone networks, including convolutional network (ConvNet) and Vision Transformer (ViT) structures, are constructed and integrated with attention modules, specifically designed to amplify the perception of insulators and defective regions. The geometric scale of anchor boxes is reconstructed using a developed clustering algorithm, considering the elongated characteristics of insulator strings to improve the adaptability of anchor boxes. Bilinear interpolation is utilized to mitigate spatial misalignment issues during the downsampling process of Region Proposal Network (RPN)-based proposals. The experimental results indicate that the improved models with the Swin Transformer (Swin-T) backbone framework achieve the mean Average Precision (mAP)@0. 5 of 88. 42% and mAP@0. 7 of 60. 52%, with a defect recall rate of 81. 94%. Additionally, the improved IoU loss function contributes to the performance of the model at higher IoU thresholds. The results of this study contribute to the further development of defect detection frameworks for power vision applications.

Details DOI

JBHI Journal 2023 Journal Article

Mind the Gap: Alleviating Local Imbalance for Unsupervised Cross-Modality Medical Image Segmentation

Zixian Su
Kai Yao
Xi Yang
Qiufeng Wang
Yuyao Yan
Jie Sun
Kaizhu Huang

Unsupervised cross-modality medical image adaptation aims to alleviate the severe domain gap between different imaging modalities without using the target domain label. A key in this campaign relies upon aligning the distributions of source and target domain. One common attempt is to enforce the global alignment between two domains, which, however, ignores the fatal local-imbalance domain gap problem, i. e. , some local features with larger domain gap are harder to transfer. Recently, some methods conduct alignment focusing on local regions to improve the efficiency of model learning. While this operation may cause a deficiency of critical information from contexts. To tackle this limitation, we propose a novel strategy to alleviate the domain gap imbalance considering the characteristics of medical images, namely Global-Local Union Alignment. Specifically, a feature-disentanglement style-transfer module first synthesizes the target-like source images to reduce the global domain gap. Then, a local feature mask is integrated to reduce the ‘inter-gap’ for local features by prioritizing those discriminative features with larger domain gap. This combination of global and local alignment can precisely localize the crucial regions in segmentation target while preserving the overall semantic consistency. We conduct a series of experiments with two cross-modality adaptation tasks, i, e. cardiac substructure and abdominal multi-organ segmentation. Experimental results indicate that our method achieves state-of-the-art performance in both tasks.

Details DOI

YNICL Journal 2023 Journal Article

Personalized estimates of morphometric similarity in multiple sclerosis and neuromyelitis optica spectrum disorders

Jie Sun
Wenjin Zhao
Yingying Xie
Fuqing Zhou
Lin Wu
Yuxin Li
Haiqing Li
Yongmei Li

Brain morphometric alterations involve multiple brain regions on progression of the disease in multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) and exhibit age-related degenerative changes during the pathological aging. Recent advance in brain morphometry as measured using MRI have leveraged Person-Based Similarity Index (PBSI) approach to assess the extent of within-diagnosis similarity or heterogeneity of brain neuroanatomical profiles between individuals of healthy populations and validate in neuropsychiatric disorders. Brain morphometric changes throughout the lifespan would be invaluable for understanding regional variability of age-related structural degeneration and the substrate of inflammatory demyelinating disease. Here, we aimed to quantify the neuroanatomical profiles with PBSI measures of cortical thickness (CT) and subcortical volumes (SV) in 263 MS, 207 NMOSD, and 338 healthy controls (HC) from six separate central datasets (aged 11-80). We explored the between-group comparisons of PBSI measures, as well as the advancing age and sex effects on PBSI measures. Compared to NMOSD, MS showed a lower extent of within-diagnosis similarity. Significant differences in regional contributions to PBSI score were observed in 29 brain regions between MS and NMOSD (P < 0.05/164, Bonferroni corrected), of which bilateral cerebellum in MS and bilateral parahippocampal gyrus in NMOSD represented the highest divergence between the two patient groups, with a high similarity effect within each group. The PBSI scores were generally lower with advancing age, but their associations showed different patterns depending on the age range. For MS, CT profiles were significantly negatively correlated with age until the early 30 s (ρ = -0.265, P = 0.030), while for NMOSD, SV profiles were significantly negatively correlated with age with 51 year-old and older (ρ = -0.365, P = 0.008). The current study suggests that PBSI approach could be used to quantify the variation in brain morphometric changes in CNS inflammatory demyelinating disease, and exhibited a greater neuroanatomical heterogeneity pattern in MS compared with NMOSD. Our results reveal that, as an MR marker, PBSI may be sensitive to distribute the disease-associated grey matter diversity and complexity. Disease-driven production of regionally selective and age stage-dependency changes in the neuroanatomical profile of MS and NMOSD should be considered to facilitate the prediction of clinical outcomes and assessment of treatment responses.

Details DOI

AAAI Conference 2023 Conference Paper

Rethinking Data Augmentation for Single-Source Domain Generalization in Medical Image Segmentation

Zixian Su
Kai Yao
Xi Yang
Kaizhu Huang
Qiufeng Wang
Jie Sun

Single-source domain generalization (SDG) in medical image segmentation is a challenging yet essential task as domain shifts are quite common among clinical image datasets. Previous attempts most conduct global-only/random augmentation. Their augmented samples are usually insufficient in diversity and informativeness, thus failing to cover the possible target domain distribution. In this paper, we rethink the data augmentation strategy for SDG in medical image segmentation. Motivated by the class-level representation invariance and style mutability of medical images, we hypothesize that unseen target data can be sampled from a linear combination of C (the class number) random variables, where each variable follows a location-scale distribution at the class level. Accordingly, data augmented can be readily made by sampling the random variables through a general form. On the empirical front, we implement such strategy with constrained Bezier transformation on both global and local (i.e. class-level) regions, which can largely increase the augmentation diversity. A Saliency-balancing Fusion mechanism is further proposed to enrich the informativeness by engaging the gradient information, guiding augmentation with proper orientation and magnitude. As an important contribution, we prove theoretically that our proposed augmentation can lead to an upper bound of the generalization risk on the unseen target domain, thus confirming our hypothesis. Combining the two strategies, our Saliency-balancing Location-scale Augmentation (SLAug) exceeds the state-of-the-art works by a large margin in two challenging SDG tasks. Code is available at https://github.com/Kaiseem/SLAug.

PDF Details DOI

JBHI Journal 2022 Journal Article

A Novel 3D Unsupervised Domain Adaptation Framework for Cross-Modality Medical Image Segmentation

Kai Yao
Zixian Su
Kaizhu Huang
Xi Yang
Jie Sun
Amir Hussain
Frans Coenen

We consider the problem of volumetric (3D) unsupervised domain adaptation (UDA) in cross-modality medical image segmentation, aiming to perform segmentation on the unannotated target domain (e. g. MRI) with the help of labeled source domain (e. g. CT). Previous UDA methods in medical image analysis usually suffer from two challenges: 1) they focus on processing and analyzing data at 2D level only, thus missing semantic information from the depth level; 2) one-to-one mapping is adopted during the style-transfer process, leading to insufficient alignment in the target domain. Different from the existing methods, in our work, we conduct a first of its kind investigation on multi-style image translation for complete image alignment to alleviate the domain shift problem, and also introduce 3D segmentation in domain adaptation tasks to maintain semantic consistency at the depth level. In particular, we develop an unsupervised domain adaptation framework incorporating a novel quartet self-attention module to efficiently enhance relationships between widely separated features in spatial regions on a higher dimension, leading to a substantial improvement in segmentation accuracy in the unlabeled target domain. In two challenging cross-modality tasks, specifically brain structures and multi-organ abdominal segmentation, our model is shown to outperform current state-of-the-art methods by a significant margin, demonstrating its potential as a benchmark resource for the biomedical and health informatics research community.

Details DOI

YNIMG Journal 2022 Journal Article

Electrophysiological correlates of confidence differ across correct and erroneous perceptual decisions

Daniel Feuerriegel
Mackenzie Murphy
Alexandra Konski
Vinay Mepani
Jie Sun
Robert Hester
Stefan Bode

Every decision we make is accompanied by an estimate of the probability that our decision is accurate or appropriate. This probability estimate is termed our degree of decision confidence. Recent work has uncovered event-related potential (ERP) correlates of confidence both during decision formation and after a decision has been made. However, the interpretation of these findings is complicated by methodological issues related to ERP amplitude measurement that are prevalent across existing studies. To more accurately characterise the neural correlates of confidence, we presented participants with a difficult perceptual decision task that elicited a broad range of confidence ratings. We identified a frontal ERP component within an onset prior to the behavioural response, which exhibited more positive-going amplitudes in trials with higher confidence ratings. This frontal effect also biased measures of the centro-parietal positivity (CPP) component at parietal electrodes via volume conduction. Amplitudes of the error positivity (Pe) component that followed each decision were negatively associated with confidence for trials with decision errors, but not for trials with correct decisions, with Bayes factors providing moderate evidence for the null in the latter case. We provide evidence for both pre- and post-decisional neural correlates of decision confidence that are observed in trials with correct and erroneous decisions, respectively. Our findings suggest that certainty in having made a correct response is associated with frontal activity during decision formation, whereas certainty in having committed an error is instead associated with the post-decisional Pe component. These findings also highlight the possibility that some previously reported associations between decision confidence and CPP/Pe component amplitudes may have been a consequence of ERP amplitude measurement-related confounds. Re-analysis of existing datasets may be useful to test this hypothesis more directly.

Details DOI

ICRA Conference 2022 Conference Paper

GCLO: Ground Constrained LiDAR Odometry with Low-drifts for GPS-denied Indoor Environments

Xin Wei
Jixin Lv
Jie Sun
Erbao Dong
Shiliang Pu

LiDAR is widely adopted in Simultaneous Localization And Mapping (SLAM) and High Definition (HD) map production. The accuracy of LiDAR Odometry (LO) is of great importance, especially in GPS-denied environments. However, we found typical LO results are prone to drift upwards along the vertical direction in underground parking lots, leading to poor mapping results. This paper proposes a Ground Constrained LO method named GCLO, which exploits planar grounds in these specific environments to compress the vertical pose drifts. GCLO is divided into three parts. First, a sensor-centric sliding map is maintained, and the point-to-plane ICP method is implemented to perform the scan-to-map registration. Then, at each key-frame, the sliding map is recorded as a local map. Ground points nearby are segmented and modeled as a planar landmark in the form of Closest Point (CP) parameterization. Finally, planar ground landmarks observed at different key-frames are associated. The ground landmark observation constraints are fused into the pose graph optimization framework to improve the LO performance. Experimental results in HIK and KITTI datasets demonstrate GCLO's superior performances in terms of accuracy in indoor multi-floor parking lots and flat outdoor sites. The limitation of GCLO in adaptability for other environments is also discussed.

Details

EAAI Journal 2022 Journal Article

Geometrically interpretable Variance Hyper Rectangle learning for pattern classification

Jie Sun
Huamao Gu
Haoyu Peng
Yili Fang
Xun Wang

Many current intrinsically interpretable machine learning models can only handle the data that are linear, low-dimensional, and relatively independent attributes and often with discrete attribute values, while the models that are capable of handling high-dimensional nonlinear data, like deep learning, have very poor interpretability. Based on the geometric characteristics, a new idea of accurately wrapping the data region with minimum-volume geometry is proposed for pattern classification. The Variance Hyper Rectangle (VHR) model presented in this paper is a realization of the idea. The VHR model uses the minimum-volume hyper rectangles, obtained through projection variance calculation, to wrap the regions occupied by a category of data, hence it has strong and clear geometric interpretability. In addition, the VHR model is well suited for large data volume, as it approaches the linear complexity in both time and space. Extensive qualitative and quantitative experiments are performed on seven real-world data sets, demonstrating that VHR outperforms the state-of-the-art interpretable methods while running quickly.

Details DOI

IJCAI Conference 2022 Conference Paper

On the Opportunity of Causal Learning in Recommendation Systems: Foundation, Estimation, Prediction and Challenges

Peng Wu
Haoxuan Li
Yuhao Deng
Wenjie Hu
Quanyu Dai
Zhenhua Dong
Jie Sun
Rui Zhang

Recently, recommender system (RS) based on causal inference has gained much attention in the industrial community, as well as the states of the art performance in many prediction and debiasing tasks. Nevertheless, a unified causal analysis framework has not been established yet. Many causal-based prediction and debiasing studies rarely discuss the causal interpretation of various biases and the rationality of the corresponding causal assumptions. In this paper, we first provide a formal causal analysis framework to survey and unify the existing causal-inspired recommendation methods, which can accommodate different scenarios in RS. Then we propose a new taxonomy and give formal causal definitions of various biases in RS from the perspective of violating the assumptions adopted in causal analysis. Finally, we formalize many debiasing and prediction tasks in RS, and summarize the statistical and machine learning-based causal estimation methods, expecting to provide new research opportunities and perspectives to the causal RS community.

PDF Details DOI

IROS Conference 2022 Conference Paper

SO-PFH: Semantic Object-based Point Feature Histogram for Global Localization in Parking Lot

Jixin Lv
Chao Meng
Yue Wang 0020
Jie Sun
Rong Xiong
Shiliang Pu

Global localization is essential for autonomous mobile systems, especially indoor applications where the GPS signal is denied. Although the appearance-based methods have been successfully applied in various localization tasks, they face various challenges such as light variation, viewpoint changing, and dynamic interference. Additionally, the appearance-based methods usually require a visual feature point map, which increases the storage burden. This paper proposes a novel global localization solution that leverages sparse and repetitive semantic object information. The proposal can fulfill global localization based on object-level maps that are self-built or externally provided. In this solution, the semantic objects are firstly modeled with a point cloud. Then, the object's semantic information is embedded into the geometry of the corresponding point, and the Semantic Object-based Point Feature Histogram (SO-PFH) descriptors of the modeled point clouds are estimated. Finally, the global localization is executed by applying a Geometric Consistency Filter-based RANdom SAmple Consensus (GCF-RANSAC) method to match point clouds. Experiments and simulations are conducted in indoor parking lots. The results demonstrate the effectiveness of the proposed method.

Details