Yongjun Xu

AAAI Conference 2026 Conference Paper

APT: Affine Prototype-Timestamp for Time Series Forecasting Under Distribution Shift

• Yujie Li
• Zezhi Shao
• Chengqing Yu
• Yisong Fu
• Tao Sun
• Yongjun Xu
• Fei Wang

Time series forecasting under distribution shift remains challenging, as existing deep learning models often rely on local statistical normalization (e.g., mean and variance) that fails to capture global distribution shift. Methods like RevIN and its variants attempt to decouple distribution and pattern but still struggle with missing values, noisy observations, and invalid channel-wise affine transformation. To address these limitations, we propose Affine Prototype-Timestamp(APT), a lightweight and flexible plug-in module that injects global distribution features into the normalization–forecasting pipeline. By leveraging timestamp-conditioned prototype learning, APT dynamically generates affine parameters that modulate both input and output series, enabling the backbone to learn from self-supervised, distribution-aware clustered instances. APT is compatible with arbitrary forecasting backbones and normalization strategies while introducing minimal computational overhead. Extensive experiments across six benchmark datasets and multiple backbone-normalization combinations demonstrate that APT significantly improves forecasting performance under distribution shift.

NeurIPS Conference 2025 Conference Paper

$\text{S}^2$Q-VDiT: Accurate Quantized Video Diffusion Transformer with Salient Data and Sparse Token Distillation

• Weilun Feng
• Haotong Qin
• Chuanguang Yang
• Xiangqi Li
• Han Yang
• Yuqi Li
• Zhulin An
• Libo Huang

Diffusion transformers have emerged as the mainstream paradigm for video generation models. However, the use of up to billions of parameters incurs significant computational costs. Quantization offers a promising solution by reducing memory usage and accelerating inference. Nonetheless, we observe that the joint modeling of spatial and temporal information in video diffusion models (V-DMs) leads to extremely long token sequences, which introduces high calibration variance and learning challenges. To address these issues, we propose **$S^2$Q-VDiT**, a post-training quantization framework for V-DMs that leverages **S**alient data and **S**parse token distillation. During the calibration phase, we identify that quantization performance is highly sensitive to the choice of calibration data. To mitigate this, we introduce *Hessian-aware Salient Data Selection*, which constructs high-quality calibration datasets by considering both diffusion and quantization characteristics unique to V-DMs. To tackle the learning challenges, we further analyze the sparse attention patterns inherent in V-DMs. Based on this observation, we propose *Attention-guided Sparse Token Distillation*, which exploits token-wise attention distributions to emphasize tokens that are more influential to the model's output. Under W4A6 quantization, $S^2$Q-VDiT achieves lossless performance while delivering $3. 9\times$ model compression and $1. 3\times$ inference acceleration. Code will be available at https: //github. com/wlfeng0509/s2q-vdit.

AAAI Conference 2025 Conference Paper

GTDE: Grouped Training with Decentralized Execution for Multi-agent Actor-Critic

• Mengxian Li
• Qi Wang
• Yongjun Xu

The rapid advancement of multi-agent reinforcement learning (MARL) has given rise to diverse training paradigms to learn the policies of each agent in the multi-agent system. The paradigms of decentralized training and execution (DTDE) and centralized training with decentralized execution (CTDE) have been proposed and widely applied. However, as the number of agents increases, the inherent limitations of these frameworks significantly degrade the performance metrics, such as win rate, total reward, etc. To reduce the influence of the increasing number of agents on the performance metrics, we propose a novel training paradigm of grouped training decentralized execution (GTDE). This framework eliminates the need for a centralized module and relies solely on local information, effectively meeting the training requirements of large-scale multi-agent systems. Specifically, we first introduce an adaptive grouping module, which divides each agent into different groups based on their observation history. To implement end-to-end training, GTDE uses Gumbel-Sigmoid for efficient point-to-point sampling on the grouping distribution while ensuring gradient backpropagation. To adapt to the uncertainty in the number of members in a group, two methods are used to implement a group information aggregation module that merges member information within the group. Empirical results show that in a cooperative environment with 495 agents, GTDE increased the total reward by an average of 382% compared to the baseline. In a competitive environment with 64 agents, GTDE achieved a 100% win rate against the baseline.

AAAI Conference 2025 Conference Paper

HSRDiff: A Hierarchical Self-Regulation Diffusion Model for Stochastic Semantic Segmentation

• Han Yang
• Chuanguang Yang
• Zhulin An
• Libo Huang
• Yongjun Xu

In safety-critical domains such as medical diagnostics and autonomous driving, single-image evidence is sometimes insufficient to reflect the inherent ambiguity of vision problems. Therefore, multiple plausible assumptions that match the image semantics may be needed to reflect the actual distribution of targets and support downstream tasks. However, balancing and improving the diversity and consistency of segmentation predictions under the high-dimensional output spaces and potential multimodal distributions is still challenging. This paper presents Hierarchical Self-Regulation Diffusion (HSRDiff), a unified framework that simulates joint probability distribution over entire labels. Our model self-regulates the balance between the two modes of predicting the label and noise in a novel ``differentiation to unification" pipeline and dynamically fits the optimal path to model the aleatoric uncertainty rooted in observations. In addition, we preserve the high-fidelity reconstruction of the delicate structure in images by leveraging the hierarchical multi-scale condition priors. We validate HSRDiff in three different semantic scenarios. Experimental results show that HSRDiff is superior to the comparison method with a considerable performance gap.

AAAI Conference 2025 Conference Paper

MPQ-DM: Mixed Precision Quantization for Extremely Low Bit Diffusion Models

• Weilun Feng
• Haotong Qin
• Chuanguang Yang
• Zhulin An
• Libo Huang
• Boyu Diao
• Fei Wang
• Renshuai Tao

Diffusion models have received wide attention in generation tasks. However, the expensive computation cost prevents the application of diffusion models in resource-constrained scenarios. Quantization emerges as a practical solution that significantly saves storage and computation by reducing the bit-width of parameters. However, the existing quantization methods for diffusion models still cause severe degradation in performance, especially under extremely low bit-widths (2-4 bit). The primary decrease in performance comes from the significant discretization of activation values at low bit quantization. Too few activation candidates are unfriendly for outlier significant weight channel quantization, and the discretized features prevent stable learning over different time steps of the diffusion model. This paper presents MPQ-DM, a Mixed-Precision Quantization method for Diffusion Models. The proposed MPQ-DM mainly relies on two techniques: (1) To mitigate the quantization error caused by outlier severe weight channels, we propose an Outlier-Driven Mixed Quantization (OMQ) technique that uses Kurtosis to quantify outlier salient channels and apply optimized intra-layer mixed-precision bit-width allocation to recover accuracy performance within target efficiency. (2) To robustly learn representations crossing time steps, we construct a Time-Smoothed Relation Distillation (TRD) scheme between the quantized diffusion model and its full-precision counterpart, transferring discrete and continuous latent to a unified relation space to reduce the representation inconsistency. Comprehensive experiments demonstrate that MPQ-DM achieves significant accuracy gains under extremely low bit-widths compared with SOTA quantization methods. MPQ-DM achieves a 58% FID decrease under W2A4 setting compared with baseline, while all other methods even collapse.

AAAI Conference 2025 Conference Paper

Multi-Teacher Knowledge Distillation with Reinforcement Learning for Visual Recognition

• Chuanguang Yang
• XinQiang Yu
• Han Yang
• Zhulin An
• Chengqing Yu
• Libo Huang
• Yongjun Xu

Multi-teacher Knowledge Distillation (KD) transfers diverse knowledge from a teacher pool to a student network. The core problem of multi-teacher KD is how to balance distillation strengths among various teachers. Most existing methods often develop weighting strategies from an individual perspective of teacher performance or teacher-student gaps, lacking comprehensive information for guidance. This paper proposes Multi-Teacher Knowledge Distillation with Reinforcement Learning (MTKD-RL) to optimize multi-teacher weights. In this framework, we construct both teacher performance and teacher-student gaps as state information to an agent. The agent outputs the teacher weight and can be updated by the return reward from the student. MTKD-RL reinforces the interaction between the student and teacher using an agent in an RL-based decision mechanism, achieving better matching capability with more meaningful weights. Experimental results on visual recognition tasks, including image classification, object detection, and semantic segmentation tasks, demonstrate that MTKD-RL achieves state-of-the-art performance compared to the existing multi-teacher KD works.

NeurIPS Conference 2025 Conference Paper

On the Integration of Spatial-Temporal Knowledge: A Lightweight Approach to Atmospheric Time Series Forecasting

• Yisong Fu
• Fei Wang
• Zezhi Shao
• Boyu Diao
• Lin Wu
• Zhulin An
• Chengqing Yu
• Yujie Li

Transformers have gained attention in atmospheric time series forecasting (ATSF) for their ability to capture global spatial-temporal correlations. However, their complex architectures lead to excessive parameter counts and extended training times, limiting their scalability to large-scale forecasting. In this paper, we revisit ATSF from a theoretical perspective of atmospheric dynamics and uncover a key insight: spatial-temporal position embedding (STPE) can inherently model spatial-temporal correlations even without attention mechanisms. Its effectiveness arises from integrating geographical coordinates and temporal features, which are intrinsically linked to atmospheric dynamics. Based on this, we propose STELLA, a S patial- T emporal knowledge E mbedded L ightweight mode L for ASTF, utilizing only STPE and an MLP architecture in place of Transformer layers. With 10k parameters and one hour of training, STELLA achieves superior performance on five datasets compared to other advanced methods. The paper emphasizes the effectiveness of spatial-temporal knowledge integration over complex architectures, providing novel insights for ATSF.

NeurIPS Conference 2025 Conference Paper

Selective Learning for Deep Time Series Forecasting

• Yisong Fu
• Zezhi Shao
• Chengqing Yu
• Yujie Li
• Zhulin An
• Qi Wang
• Yongjun Xu
• Fei Wang

Benefiting from high capacity for capturing complex temporal patterns, deep learning (DL) has significantly advanced time series forecasting (TSF). However, deep models tend to suffer from severe overfitting due to the inherent vulnerability of time series to noise and anomalies. The prevailing DL paradigm uniformly optimizes all timesteps through the MSE loss and learns those uncertain and anomalous timesteps without difference, ultimately resulting in overfitting. To address this, we propose a novel selective learning strategy for deep TSF. Specifically, selective learning screens a subset of the whole timesteps to calculate the MSE loss in optimization, guiding the model to focus on generalizable timesteps while disregarding non-generalizable ones. Our framework introduces a dual-mask mechanism to target timesteps: (1) an uncertainty mask leveraging residual entropy to filter uncertain timesteps, and (2) an anomaly mask employing residual lower bound estimation to exclude anomalous timesteps. Extensive experiments across eight real-world datasets demonstrate that selective learning can significantly improve the predictive performance for typical state-of-the-art deep models, including 37. 4% MSE reduction for Informer, 8. 4% for TimesNet, and 6. 5% for iTransformer.

NeurIPS Conference 2025 Conference Paper

SMARTraj$^2$: A Stable Multi-City Adaptive Method for Multi-View Spatio-Temporal Trajectory Representation Learning

• Tangwen Qian
• Junhe Li
• Yile Chen
• Gao Cong
• Zezhi Shao
• Jun Zhang
• Tao Sun
• Fei Wang

Spatio-temporal trajectory representation learning plays a crucial role in various urban applications such as transportation systems, urban planning, and environmental monitoring. Existing methods can be divided into single-view and multi-view approaches, with the latter offering richer representations by integrating multiple sources of spatio-temporal data. However, these methods often struggle to generalize across diverse urban scenes due to multi-city structural heterogeneity, which arises from the disparities in road networks, grid layouts, and traffic regulations across cities, and the amplified seesaw phenomenon, where optimizing for one city, view, or task can degrade performance in others. These challenges hinder the deployment of trajectory learning models across multiple cities, limiting their real-world applicability. In this work, we propose SMARTraj$^2$, a novel stable multi-city adaptive method for multi-view spatio-temporal trajectory representation learning. Specifically, we introduce a feature disentanglement module to separate domain-invariant and domain-specific features, and a personalized gating mechanism to dynamically stabilize the contributions of different views and tasks. Our approach achieves superior generalization across heterogeneous urban scenes while maintaining robust performance across multiple downstream tasks. Extensive experiments on benchmark datasets demonstrate the effectiveness of SMARTraj$^2$ in enhancing cross-city generalization and outperforming state-of-the-art methods. See our project website at \url{https: //github. com/GestaltCogTeam/SMARTraj}.

NeurIPS Conference 2024 Conference Paper

Continual Learning in the Frequency Domain

• Ruiqi Liu
• Boyu Diao
• Libo Huang
• Zijia An
• Zhulin An
• Yongjun Xu

Continual learning (CL) is designed to learn new tasks while preserving existing knowledge. Replaying samples from earlier tasks has proven to be an effective method to mitigate the forgetting of previously acquired knowledge. However, the current research on the training efficiency of rehearsal-based methods is insufficient, which limits the practical application of CL systems in resource-limited scenarios. The human visual system (HVS) exhibits varying sensitivities to different frequency components, enabling the efficient elimination of visually redundant information. Inspired by HVS, we propose a novel framework called Continual Learning in the Frequency Domain (CLFD). To our knowledge, this is the first study to utilize frequency domain features to enhance the performance and efficiency of CL training on edge devices. For the input features of the feature extractor, CLFD employs wavelet transform to map the original input image into the frequency domain, thereby effectively reducing the size of input feature maps. Regarding the output features of the feature extractor, CLFD selectively utilizes output features for distinct classes for classification, thereby balancing the reusability and interference of output features based on the frequency domain similarity of the classes across various tasks. Optimizing only the input and output features of the feature extractor allows for seamless integration of CLFD with various rehearsal-based methods. Extensive experiments conducted in both cloud and edge environments demonstrate that CLFD consistently improves the performance of state-of-the-art (SOTA) methods in both precision and training efficiency. Specifically, CLFD can increase the accuracy of the SOTA CL method by up to 6. 83% and reduce the training time by 2. 6×.

EAAI Journal 2024 Journal Article

Detection method for weld defects in time-of-flight diffraction images based on multi-image fusion and feature hybrid enhancement

• Deyan Yang
• Hongquan Jiang
• Song Ai
• Tianlun Yang
• Zelin Zhi
• Deqiang Jing
• Jianmin Gao
• Kun Yue

AAAI Conference 2024 Conference Paper

eTag: Class-Incremental Learning via Embedding Distillation and Task-Oriented Generation

• Libo Huang
• Yan Zeng
• Chuanguang Yang
• Zhulin An
• Boyu Diao
• Yongjun Xu

Class incremental learning (CIL) aims to solve the notorious forgetting problem, which refers to the fact that once the network is updated on a new task, its performance on previously-learned tasks degenerates catastrophically. Most successful CIL methods store exemplars (samples of learned tasks) to train a feature extractor incrementally, or store prototypes (features of learned tasks) to estimate the incremental feature distribution. However, the stored exemplars would violate the data privacy concerns, while the fixed prototypes might not reasonably be consistent with the incremental feature distribution, hindering the exploration of real-world CIL applications. In this paper, we propose a data-free CIL method with embedding distillation and Task-oriented generation (eTag), which requires neither exemplar nor prototype. Embedding distillation prevents the feature extractor from forgetting by distilling the outputs from the networks' intermediate blocks. Task-oriented generation enables a lightweight generator to produce dynamic features, fitting the needs of the top incremental classifier. Experimental results confirm that the proposed eTag considerably outperforms state-of-the-art methods on several benchmark datasets.

AAAI Conference 2022 Conference Paper

Interpretable Generative Adversarial Networks

• Chao Li
• Kelu Yao
• Jin Wang
• Boyu Diao
• Yongjun Xu
• Quanshi Zhang

Learning a disentangled representation is still a challenge in the field of the interpretability of generative adversarial networks (GANs). This paper proposes a generic method to modify a traditional GAN into an interpretable GAN, which ensures that filters in an intermediate layer of the generator encode disentangled localized visual concepts. Each filter in the layer is supposed to consistently generate image regions corresponding to the same visual concept when generating different images. The interpretable GAN learns to automatically discover meaningful visual concepts without any annotations of visual concepts. The interpretable GAN enables people to modify a specific visual concept on generated images by manipulating feature maps of the corresponding filters in the layer. Our method can be broadly applied to different types of GANs. Experiments have demonstrated the effectiveness of our method.

AAAI Conference 2022 Conference Paper

Mutual Contrastive Learning for Visual Representation Learning

• Chuanguang Yang
• Zhulin An
• Linhang Cai
• Yongjun Xu

We present a collaborative learning method called Mutual Contrastive Learning (MCL) for general visual representation learning. The core idea of MCL is to perform mutual interaction and transfer of contrastive distributions among a cohort of networks. A crucial component of MCL is Interactive Contrastive Learning (ICL). Compared with vanilla contrastive learning, ICL can aggregate cross-network embedding information and maximize the lower bound to the mutual information between two networks. This enables each network to learn extra contrastive knowledge from others, leading to better feature representations for visual recognition tasks. We emphasize that the resulting MCL is conceptually simple yet empirically powerful. It is a generic framework that can be applied to both supervised and self-supervised representation learning. Experimental results on image classification and transfer learning to object detection show that MCL can lead to consistent performance gains, demonstrating that MCL can guide the network to generate better feature representations. Code is available at https: //github. com/winycg/MCL.

AAAI Conference 2022 Conference Paper

Prior Gradient Mask Guided Pruning-Aware Fine-Tuning

• Linhang Cai
• Zhulin An
• Chuanguang Yang
• Yangchun Yan
• Yongjun Xu

We proposed a Prior Gradient Mask Guided Pruning-Aware Fine-Tuning (PGMPF) framework to accelerate deep Convolutional Neural Networks (CNNs). In detail, the proposed PGMPF selectively suppresses the gradient of those ”unimportant” parameters via a prior gradient mask generated by the pruning criterion during fine-tuning. PGMPF has three charming characteristics over previous works: (1) Pruningaware network fine-tuning. A typical pruning pipeline consists of training, pruning and fine-tuning, which are relatively independent, while PGMPF utilizes a variant of the pruning mask as a prior gradient mask to guide fine-tuning, without complicated pruning criteria. (2) An excellent tradeoff between large model capacity during fine-tuning and stable convergence speed to obtain the final compact model. Previous works preserve more training information of pruned parameters during fine-tuning to pursue better performance, which would incur catastrophic non-convergence of the pruned model for relatively large pruning rates, while our PGMPF greatly stabilizes the fine-tuning phase by gradually constraining the learning rate of those ”unimportant” parameters. (3) Channel-wise random dropout of the prior gradient mask to impose some gradient noise to fine-tuning to further improve the robustness of final compact model. Experimental results on three image classification benchmarks CI- FAR10/100 and ILSVRC-2012 demonstrate the effectiveness of our method for various CNN architectures, datasets and pruning rates. Notably, on ILSVRC-2012, PGMPF reduces 53. 5% FLOPs on ResNet-50 with only 0. 90% top-1 accuracy drop and 0. 52% top-5 accuracy drop, which has advanced the state-of-the-art with negligible extra computational cost.

IJCAI Conference 2021 Conference Paper

Hierarchical Self-supervised Augmented Knowledge Distillation

• Chuanguang Yang
• Zhulin An
• Linhang Cai
• Yongjun Xu

Knowledge distillation often involves how to define and transfer knowledge from teacher to student effectively. Although recent self-supervised contrastive knowledge achieves the best performance, forcing the network to learn such knowledge may damage the representation learning of the original class recognition task. We therefore adopt an alternative self-supervised augmented task to guide the network to learn the joint distribution of the original recognition task and self-supervised auxiliary task. It is demonstrated as a richer knowledge to improve the representation power without losing the normal classification capability. Moreover, it is incomplete that previous methods only transfer the probabilistic knowledge between the final layers. We propose to append several auxiliary classifiers to hierarchical intermediate feature maps to generate diverse self-supervised knowledge and perform the one-to-one transfer to teach the student network thoroughly. Our method significantly surpasses the previous SOTA SSKD with an average improvement of 2. 56% on CIFAR-100 and an improvement of 0. 77% on ImageNet across widely used network pairs. Codes are available at https: //github. com/winycg/HSAKD.

AAAI Conference 2020 Conference Paper

Gated Convolutional Networks with Hybrid Connectivity for Image Classification

• Chuanguang Yang
• Zhulin An
• Hui Zhu
• Xiaolong Hu
• Kun Zhang
• Kaiqiang Xu
• Chao Li
• Yongjun Xu

We propose a simple yet effective method to reduce the redundancy of DenseNet by substantially decreasing the number of stacked modules by replacing the original bottleneck by our SMG module, which is augmented by local residual. Furthermore, SMG module is equipped with an efficient twostage pipeline, which aims to DenseNet-like architectures that need to integrate all previous outputs, i. e. , squeezing the incoming informative but redundant features gradually by hierarchical convolutions as a hourglass shape and then exciting it by multi-kernel depthwise convolutions, the output of which would be compact and hold more informative multi-scale features. We further develop a forget and an update gate by introducing the popular attention modules to implement the effective fusion instead of a simple addition between reused and new features. Due to the Hybrid Connectivity (nested combination of global dense and local residual) and Gated mechanisms, we called our network as the HCGNet. Experimental results on CIFAR and ImageNet datasets show that HCGNet is more prominently efficient than DenseNet, and can also significantly outperform state-of-the-art networks with less complexity. Moreover, HCGNet also shows the remarkable interpretability and robustness by network dissection and adversarial defense, respectively. On MS-COCO, HCGNet can consistently learn better features than popular backbones.