Author name cluster

Shiyu Liu

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.

12 papers

2 author rows

AAAI Conference 2026 Conference Paper

AirDDE: Multifactor Neural Delay Differential Equations for Air Quality Forecasting

Binqing Wu
Zongjiang Shang
Shiyu Liu
Jianlong Huang
Jiahui Xu
Ling Chen

Accurate air quality forecasting is essential for public health and environmental sustainability, but remains challenging due to the complex pollutant dynamics. Existing deep learning methods often model pollutant dynamics as an instantaneous process, overlooking the intrinsic delays in pollutant propagation. Thus, we propose AirDDE, the first neural delay differential equation framework in this task that integrates delay modeling into a continuous-time pollutant evolution under physical guidance. Specifically, two novel components are introduced: (1) a memory-augmented attention module that retrieves globally and locally historical features, which can adaptively capture delay effects modulated by multifactor data; and (2) a physics-guided delay evolving function, grounded in the diffusion-advection equation, that models diffusion, delayed advection, and source/sink terms, which can capture delay-aware pollutant accumulation patterns with physical plausibility. Extensive experiments on three real-world datasets demonstrate that AirDDE achieves the state-of-the-art forecasting performance with an average MAE reduction of 8.79% over the best baselines.

PDF Details DOI

AAAI Conference 2026 Conference Paper

MoCast: Learning Turbulent Motions Under Physical Guidance for Precipitation Nowcasting

Binqing Wu
Weiqi Chen
Shiyu Liu
Zongjiang Shang
Haiou Wang
Liang Sun
Ling Chen

Precipitation nowcasting, a critical task for weather-sensitive applications, is highly challenging owing to the chaotic nature of atmospheric dynamics. Despite recent progress in deep learning, existing methods are limited in their capacity to model turbulent motions, one of the key drivers of precipitation evolution. Thus, we propose MoCast, the first work that incorporates turbulence knowledge to decompose turbulent motions into solvable components for precipitation nowcasting. Specifically, inspired by the continuity equation, MoCast introduces two core innovations: (1) a physics-guided motion module that learns turbulent motions from physically interpretable mean and fluctuating components based on Reynolds, Helmholtz, and Wavelet decomposition techniques, and (2) a motion-guided source-sink module that learns source-sink features considering the multi-scale impact from motions based on a mixture-of-experts architecture. Extensive experiments on three real-world datasets demonstrate that MoCast achieves the state-of-the-art performance. MoCast and its diffusion-based variant MoCast+ reduce CSI error by an average of 4.9% and 4.5% compared to the best deterministic and probabilistic baselines, respectively.

PDF Details DOI

JBHI Journal 2025 Journal Article

An Improved Microbial Object Detection Method for Low-Contrast and Occluded Scenarios Based on SMA-YOLOv8s

Kun Sun
Zhenqiang Song
Jiaxing Zhang
Shiyu Liu
Yu Zhang
Qinghao Song
Qing Wu

Accurate detection and localization of microbial targets are critical for microbial trajectory tracking and analysis. However, microscopic microorganism images often exhibit low contrast and mutual occlusion between targets, which pose significant challenges for microbial object accuracy detection due to insufficient distinguishable shallow-layer information and occluded targets inadequate representation. To address these issues, a novel method of SMA-YOLOv8s is proposed for microbial object detection. Firstly, the traditional strided convolutions is replaced with SPD-Conv in downsampling of YOLOv8s to retain shallow-layer information. Secondly, a feature fusion strategy that integrates Cascaded Group Attention with Scale Sequence Feature Fusion (CSFF) is proposed, which could enrich contextual feature representation for better detecting occluded targets. Thirdly, the Wise-IoU loss function is employed to optimize bounding box regression, improving localization precision. Experimental evaluations on the BCCD, CTMCv1, and a self-constructed microscopic microorganism dataset demonstrate that SMA-YOLOv8s achieves mAP50 scores of 95. 5%, 90. 3%, and 81. 7%, respectively, surpassing baseline methods in overall performance. These results highlight the robustness and effectiveness of the proposed method in detecting microbial targets under low contrast and occlusion conditions.

Details DOI

AAAI Conference 2025 Conference Paper

Efficient Deformable Convolutional Prompt for Continual Test-Time Adaptation in Medical Image Segmentation

Shiyu Liu
Daoqiang Zhang
Xiaoke Hao

The domain gap resulting from mismatches in acquisition details like protocol and scanner between training and test data hinders the deployment of the trained model in clinical practice. To address this issue, Continual test-time adaptation (CTTA) has been proposed to adapt the source model to continually changing unlabeled domains without accessing the source data. Existing methods learn an image-level visual prompt for target domains and inject the trainable prompt into the input space. However, they either combine the input with a prompt of equal scale or determine the prompt injection position through complex strategies such as uncertainty estimation or Fourier Transform. These approaches substantially increase the number of trainable parameters and computational burden, especially in high-dimensional medical imaging data. To overcome these challenges, we propose the Efficient Deformable Convolutional Prompt (EDCP), which leverages the inductive bias of convolution to reduce trainable parameters compared to standard prompts. We further enhance convolution by making it deformable, addressing fine-grained domain shifts at the pixel level through an offset branch. To improve training efficiency and balance parameters between the convolution and offset branches, we decompose the offset transformation into two parts, storing one in an offset bank that also serves as a domain indicator. This bank accelerates training by skipping test images similar to those already stored. Prompt updates are guided by layer-wise alignment of source-target statistics without unfreezing batch normalization layers. Extensive experiments demonstrate the superiority of our method in 2D and 3D medical image segmentation tasks.

PDF Details DOI

JAIR Journal 2025 Journal Article

Improving Mutual Information Based Feature Selection by Boosting Unique Relevance

Shiyu Liu
Mehul Motani

Mutual Information (MI) based feature selection makes use of MI to evaluate each feature and eventually shortlists a relevant feature subset, in order to address issues associated with high-dimensional datasets. Despite the effectiveness of MI in feature selection, we notice that many state-of-the-art algorithms disregard the so-called unique relevance (UR) of features, which is a necessary condition for the optimal feature subset. In our study of five representative MI based feature selection (MIBFS) algorithms, we find that all of them underperform as they ignore the UR of features and arrive at a suboptimal selected feature subset. We point out that the heart of the problem is that all these MIBFS algorithms follow the criterion of Maximize Relevance with Minimum Redundancy (MRwMR), which does not explicitly target UR. This motivates us to augment the existing criterion with the objective of boosting unique relevance (BUR), leading to a new criterion called MRwMR-BUR. Depending on the task being addressed, MRwMR-BUR has two variants, termed MRwMR-BUR-KSG and MRwMR-BUR-CLF, which estimate UR differently. MRwMR-BUR-KSG estimates UR via a nearest-neighbor based approach called the KSG estimator and is designed for three major tasks: (i) Classification Performance (i.e., higher classification accuracy). (ii) Feature Interpretability (i.e., a more precise selected feature subset for practitioners to explore the hidden relationship between features and labels). (iii) Classifier Generalization (i.e., the selected feature subset generalizes well to various classifiers). MRwMR-BUR-CLF estimates UR via a classifier based approach. It adapts UR to different classifiers, further improving the competitiveness of MRwMR-BUR for classification performance oriented tasks. The performance of MRwMR-BUR-KSG and MRwMR-BUR-CLF is validated via experiments using six public datasets and four popular classifiers. Specifically, as compared to MRwMR, the proposed MRwMR-BUR-KSG improves the test accuracy by 2% – 3% with 25% – 30% fewer features being selected, without increasing the algorithm complexity. MRwMR-BUR-CLF further improves the classification performance by 3.8% – 5.5% (relative to MRwMR), and it also outperforms three popular classifier dependent feature selection methods.

PDF Details DOI

JBHI Journal 2025 Journal Article

Medical Hyperspectral Image Feature Selection Framework Using Functional Data Analysis: Application to Membranous Nephropathy Pathologic Diagnosis

Meng Lv
Shiyu Liu
Xiaoying Ma
Yue Yang
Haihao Zhang
Wei Li

To address the core issue of high-dimensional data processing in hyperspectral pathological diagnosis, we develop a new feature selection framework using functional data analysis (FSFDA). The framework models pixel spectra as continuous functions to preserve spectral continuity, overcoming the limitations of traditional discrete representations. Based on functional features, an innovative adaptive spectral segmentation strategy driven by functional change rate is developed to achieve optimal segmentation in the feature space. Additionally, a multi-criteria scoring mechanism including supervised (FSFDA-S) and unsupervised (FSFDA-U) paradigms is developed to enhance feature diagnostic discriminability while maintaining sparsity. Experimental results on the pathological hyperspectral image dataset of membranous nephropathy validate that the proposed method achieves over 99% classification accuracy while reducing feature dimensions by 94. 5%. For cross-modal data involving in-vivo human brain and white blood cells, FSFDA effectively identifies diagnostic bands aligned with histopathological signatures, verifying its adaptive feature selection ability and cross sample generalization performance.

Details DOI

IJCAI Conference 2024 Conference Paper

Meta-Learning via PAC-Bayesian with Data-Dependent Prior: Generalization Bounds from Local Entropy

Shiyu Liu
Wei Shi
Zenglin Xu
Shaogao Lv
Yehong Zhang
Hui Wang

Meta-learning accelerates the learning process on unseen learning tasks by acquiring prior knowledge through previous related tasks. The PAC-Bayesian theory provides a theoretical framework to analyze the generalization of meta-learning to unseen tasks. However, previous works still encounter two notable limitations: (1) they merely focus on the data-free priors, which often result in inappropriate regularization and loose generalization bounds; (2) more importantly, their optimization process usually involves nested optimization problems, incurring significant computational costs. To address these issues, we derive new generalization bounds and introduce a novel PAC-Bayesian framework for meta-learning that integrates data-dependent priors. This framework enables the extraction of optimal posteriors for each task in closed form, thereby allowing us to minimize generalization bounds incorporated data-dependent priors with only a simple local entropy. The resulting algorithm, which employs SGLD for sampling from the optimal posteriors, is stable, efficient, and computationally lightweight, eliminating the need for nested optimization. Extensive experimental results demonstrate that our proposed method outperforms the other baselines.

PDF Details DOI

TMLR Journal 2023 Journal Article

AP: Selective Activation for De-sparsifying Pruned Networks

Shiyu Liu
Rohan Ghosh
Mehul Motani

The rectified linear unit (ReLU) is a highly successful activation function in neural networks as it allows networks to easily obtain sparse representations, which reduces overfitting in overparameterized networks. However, in the context of network pruning, we find that the sparsity introduced by ReLU, which we quantify by a term called dynamic dead neuron rate (DNR), is not beneficial for the pruned network. Interestingly, the more the network is pruned, the smaller the dynamic DNR becomes during and after optimization. This motivates us to propose a method to explicitly reduce the dynamic DNR for the pruned network, i.e., de-sparsify the network. We refer to our method as Activate-while-Pruning (AP). We note that AP does not function as a stand-alone method, as it does not evaluate the importance of weights. Instead, it works in tandem with existing pruning methods and aims to improve their performance by selective activation of nodes to reduce the dynamic DNR. We conduct extensive experiments using various popular networks (e.g., ResNet, VGG, DenseNet, MobileNet) via two classical and three state-of-the-art pruning methods. The experimental results on public datasets (e.g., CIFAR-10, CIFAR-100) suggest that AP works well with existing pruning methods and improves the performance by 3% - 4%. For larger scale datasets (e.g., ImageNet) and state-of-the-art networks (e.g., vision transformer), we observe an improvement of 2% - 3% with AP as opposed to without. Lastly, we conduct an ablation study to examine the effectiveness of the components comprising AP.

PDF Details

TMLR Journal 2023 Journal Article

Optimizing Learning Rate Schedules for Iterative Pruning of Deep Neural Networks

Shiyu Liu
Rohan Ghosh
John Chong Min Tan
Mehul Motani

The importance of learning rate (LR) schedules on network pruning has been observed in a few recent works. As an example, Frankle and Carbin (2019) highlighted that winning tickets (i.e., accuracy preserving subnetworks) can not be found without applying a LR warmup schedule. Renda, Frankle and Carbin (2020) also demonstrated that rewinding the LR to its initial state at the end of each pruning cycle can improve pruning performance. In this paper, we go one step further by first providing a theoretical justification for the surprising effect of LR schedules. Next, we propose a LR schedule for network pruning called SILO, which stands for S-shaped Improved Learning rate Optimization. The advantages of SILO over existing LR schedules are two-fold: (i) SILO has a strong theoretical motivation and dynamically adjusts the LR during pruning to improve generalization. Specifically, SILO increases the LR upper bound (max_lr) in an S-shape. This leads to an improvement of 2% - 4% in extensive experiments with various types of networks (e.g., Vision Transformers, ResNet) on popular datasets such as ImageNet, CIFAR-10/100. (ii) In addition to the strong theoretical motivation, SILO is empirically optimal in the sense of matching an Oracle, which exhaustively searches for the optimal value of max_lr via grid search. We find that SILO is able to precisely adjust the value of max_lr to be within the Oracle optimized interval, resulting in performance competitive with the Oracle with significantly lower complexity.

PDF Details

IJCAI Conference 2023 Conference Paper

Stability and Generalization of lp-Regularized Stochastic Learning for GCN

Shiyu Liu
Linsen Wei
Shaogao Lv
Ming Li

Graph convolutional networks (GCN) are viewed as one of the most popular representations among the variants of graph neural networks over graph data and have shown powerful performance in empirical experiments. That l2-based graph smoothing enforces the global smoothness of GCN, while (soft) l1-based sparse graph learning tends to promote signal sparsity to trade for discontinuity. This paper aims to quantify the trade-off of GCN between smoothness and sparsity, with the help of a general lp-regularized (1<p<= 2) stochastic learning proposed within. While stability-based generalization analyses have been given in prior work for a second derivative objectiveness function, our lp-regularized learning scheme does not satisfy such a smooth condition. To tackle this issue, we propose a novel SGD proximal algorithm for GCNs with an inexact operator. For a single-layer GCN, we establish an explicit theoretical understanding of GCN with the lp-regularized stochastic learning by analyzing the stability of our SGD proximal algorithm. We conduct multiple empirical experiments to validate our theoretical findings.

PDF Details DOI

NeurIPS Conference 2021 Conference Paper

ByPE-VAE: Bayesian Pseudocoresets Exemplar VAE

Qingzhong Ai
Lirong He
Shiyu Liu
Zenglin Xu

Recent studies show that advanced priors play a major role in deep generative models. Exemplar VAE, as a variant of VAE with an exemplar-based prior, has achieved impressive results. However, due to the nature of model design, an exemplar-based model usually requires vast amounts of data to participate in training, which leads to huge computational complexity. To address this issue, we propose Bayesian Pseudocoresets Exemplar VAE (ByPE-VAE), a new variant of VAE with a prior based on Bayesian pseudocoreset. The proposed prior is conditioned on a small-scale pseudocoreset rather than the whole dataset for reducing the computational cost and avoiding overfitting. Simultaneously, we obtain the optimal pseudocoreset via a stochastic optimization algorithm during VAE training aiming to minimize the Kullback-Leibler divergence between the prior based on the pseudocoreset and that based on the whole dataset. Experimental results show that ByPE-VAE can achieve competitive improvements over the state-of-the-art VAEs in the tasks of density estimation, representation learning, and generative data augmentation. Particularly, on a basic VAE architecture, ByPE-VAE is up to 3 times faster than Exemplar VAE while almost holding the performance. Code is available at \url{https: //github. com/Aiqz/ByPE-VAE}.

PDF Details

IROS Conference 2021 Conference Paper

Decentralized Control and Teleoperation of a Multi-UAV Parallel Robot Based on Intrinsic Measurements

Shiyu Liu
Julian Erskine
Abdelhamid Chriette
Isabelle Fantoni

Aerial manipulators have great potential in accomplishing a variety of aerial tasks. One class of aerial manipulators, multi-UAV parallel robots, consists of multiple UAVs connected to a payload or an end-effector by passive kinematic chains. The primary limitation of such aerial manipulators is the dependence on motion capture (MOCAP) systems that provide precise and high-rate exteroceptive pose measurements of all bodies in a common inertial frame, but which are impractical in the majority of real applications. This paper proposes a novel methodology of controlling multi-UAV parallel robots, using a Flying Parallel Robot (FPR) as a case study, that could be deployed without a system of external localisation. Intrinsic measurements acquired onboard the UAVs are used to recover a set of robot states that avoid using coordinates derived from a global frame and allow control of the robot by teleoperation. Two decentralized control methods are proposed, based on inter-UAV communicating or non-communicating scenarios. Experiments with intrinsic measurements emulated by MOCAP are carried out to show the performance of the proposed method.

Details