Weakly-supervised diffusion models in anomaly segmen- tation, which leverage image-level labels and bypass the need for pixel-level labels during training, have shown superior performance over unsupervised methods, offer- ing a cost-effective alternative. Traditional methods that rely on iterative image reconstruction are not fully weakly- supervised due to their dependence on costly pixel-level la- bels for hyperparameter tuning in inference. To address this issue, we introduce Anomaly Detection with Forward Process of Diffusion Models (AnoFPDM), a fully weakly- supervised framework that operates without image recon- struction and eliminates the need for pixel-level labels in hyperparameter tuning. By using the unguided forward pro- cess as a reference, AnoFPDM dynamically selects hyper- parameters such as noise scale and segmentation threshold for each input. We improve anomaly segmentation by ag- gregating anomaly maps from each step of the guided for- ward process, which strengthens the signal of anomalous regions in the aggregated anomaly map. Our framework demonstrates competitive performance on the BraTS21 and ATLAS v2.0 datasets.
@article{che2024anofpdm,title={AnoFPDM: anomaly segmentation with forward process of diffusion models for brain MRI},author={Che, Yiming and Rafsani, Fazle and Shah, Jay and Siddiquee, Md Mahfuzur Rahman and Wu, Teresa},journal={arXiv preprint arXiv:2404.15683},year={2024},}
Enhancing amyloid PET quantification: MRI-guided super-resolution using latent diffusion models
Jay Shah, Yiming Che, Javad Sohankar, and 5 more authors
@article{shah2024enhancinh,title={Enhancing amyloid PET quantification: MRI-guided super-resolution using latent diffusion models},author={Shah, Jay and Che, Yiming and Sohankar, Javad and Luo, Ji and Li, Baoxin and Su, Yi and Wu, Teresa and Initiative, Alzheimer’s Disease Neuroimaging},journal={Life},volume={14},number={12},pages={1580},year={2024},publisher={MDPI},}
@article{wan2024sparse,title={Sparse bayesian learning for sequential inference of network connectivity from small data},author={Wan, Jinming and Kataoka, Jun and Sivakumar, Jayanth and Pe{\~n}a, Eric and Che, Yiming and Sayama, Hiroki and Cheng, Changqing},journal={IEEE Transactions on Network Science and Engineering},year={2024},publisher={IEEE},}
Dispersion-enhanced sequential batch sampling for adaptive contour estimation
Yiming Che, Juliane Müller, and Changqing Cheng
Quality and Reliability Engineering International, 2024
@article{che2023physical,title={Physical-statistical learning in resilience assessment for power generation systems},author={Che, Yiming and Zhang, Ziang John and Cheng, Changqing},journal={Physica A: Statistical Mechanics and its Applications},pages={128584},year={2023},doi={10.1103/PhysRev.47.777},publisher={Elsevier}}
Uncertainty quantification and optimal robust design for machining operations
Jinming Wan, Yiming Che, Zimo Wang, and 1 more author
Journal of Computing and Information Science in Engineering, 2023
@article{wan2023uncertainty,title={Uncertainty quantification and optimal robust design for machining operations},author={Wan, Jinming and Che, Yiming and Wang, Zimo and Cheng, Changqing},journal={Journal of Computing and Information Science in Engineering},volume={23},number={1},pages={011005},year={2023},publisher={American Society of Mechanical Engineers},}
Characterizations and optimization for resilient manufacturing systems with considerations of process uncertainties
Qiyang Ma, Yiming Che, Changqing Cheng, and 1 more author
Journal of Computing and Information Science in Engineering, 2023
@article{ma2023characterizations,title={Characterizations and optimization for resilient manufacturing systems with considerations of process uncertainties},author={Ma, Qiyang and Che, Yiming and Cheng, Changqing and Wang, Zimo},journal={Journal of Computing and Information Science in Engineering},volume={23},number={1},pages={011007},year={2023},publisher={American Society of Mechanical Engineers},}
Stochastic kriging (SK) offers an explicit way to characterize heterogeneous noise variance in stochastic computer simulations and has gained considerable traction recently as a surrogate model. Nevertheless, SK relies on tedious Monte Carlo (MC) method to estimate the intrinsic variance at each design input. For computationally expensive simulations, the substantial replication effort has essentially rendered SK intractable. To this end, we develop generalized polynomial chaos (gPC)-informed efficient stochastic kriging (gPC-SK) to ameliorate the computational cost. At its core, gPC supplants the tedious repetitive MC simulations, instead resting on a much smaller set of sampling points to estimate the intrinsic uncertainty, thus applicable to those prohibitively expensive simulations. We present the gPC-SK in sequential optimal design on the borehole function and stability of time-delay dynamic systems.
@article{che2021generalized,title={Generalized polynomial chaos-informed efficient stochastic Kriging},author={Che, Yiming and Guo, Ziqi and Cheng, Changqing},journal={Journal of Computational Physics},volume={445},pages={110598},year={2021},publisher={Elsevier},}
@article{che2021active,title={Active learning and relevance vector machine in efficient estimate of basin stability for large-scale dynamic networks},author={Che, Yiming and Cheng, Changqing},journal={Chaos: An Interdisciplinary Journal of Nonlinear Science},volume={31},number={5},pages={053129},year={2021},publisher={AIP Publishing LLC}}
2020
Pattern recognition and automatic identification of early-stage atrial fibrillation
Xiaodan Wu, Yumeng Zheng, Yiming Che, and 1 more author
@article{wu2020pattern,title={Pattern recognition and automatic identification of early-stage atrial fibrillation},author={Wu, Xiaodan and Zheng, Yumeng and Che, Yiming and Cheng, Changqing},journal={Expert Systems with Applications},volume={158},pages={113560},year={2020},publisher={Elsevier}}
@article{che2020fast,title={Fast basin stability estimation for dynamic systems under large perturbations with sequential support vector machine},author={Che, Yiming and Cheng, Changqing and Liu, Zhao and Zhang, Ziang John},journal={Physica D: Nonlinear Phenomena},volume={405},pages={132381},year={2020},publisher={Elsevier}}
@article{che2019multi,title={Multi-fidelity modeling in sequential design for stability identification in dynamic time-delay systems},author={Che, Yiming and Liu, Jiachen and Cheng, Changqing},journal={Chaos: An Interdisciplinary Journal of Nonlinear Science},volume={29},number={9},pages={093105},year={2019},publisher={AIP Publishing LLC}}
Time delay is ubiquitous in many real-world physical and biological systems. It typically gives rise to rich dynamic behaviors, from aperiodic to chaotic. The stability of such dynamic behaviors is of considerable interest for process control purposes. While stability analysis under deterministic conditions has been extensively studied, not too many works addressed the issue of stability under uncertainty. Nonetheless, uncertainty, in either modeling or parameter estimation, is inevitable in complex system studies. Even for high-fidelity models, the uncertainty of input parameters could lead to divergent behaviors compared to the deterministic study. This is especially true when the system is at or near the bifurcation point. To this end, we investigated generalized polynomial chaos (GPC) to quantify the impact of uncertain parameters on the stability of delay systems. Our studies suggested that uncertainty quantification in delay systems provides richer information for system stability compared to deterministic analysis. In contrast to the robust yet time-consuming Monte Carlo or Latin hypercube sampling method, GPC approach achieves the same accuracy but only with a fraction of the computational overhead.
@article{che2018uncertainty,title={Uncertainty quantification in stability analysis of chaotic systems with discrete delays},author={Che, Yiming and Cheng, Changqing},journal={Chaos, Solitons \& Fractals},volume={116},pages={208--214},year={2018},publisher={Elsevier},}