Author name cluster

Liang Hou

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.

10 papers

2 author rows

AAAI Conference 2026 Conference Paper

Boosting Resolution Generalization of Diffusion Transformers with Randomized Positional Encodings

Liang Hou
Cong Liu
Mingwu Zheng
Xin Tao
Pengfei Wan
Di Zhang
Kun Gai

Resolution generalization in image generation tasks enables the production of higher-resolution images with lower training resolution overhead. However, a key obstacle for diffusion transformers in addressing this problem is the mismatch between positional encodings seen at inference and those used during training. Existing strategies such as positional encodings interpolation, extrapolation, or hybrids, do not fully resolve this mismatch. In this paper, we propose a novel two-dimensional randomized positional encodings, namely RPE-2D, that prioritizes the order of image patches rather than their absolute distances, enabling seamless high- and low-resolution generation without training on multiple resolutions. Concretely, RPE-2D independently samples positions along the horizontal and vertical axes over an expanded range during training, ensuring that the encodings used at inference lie within the training distribution and thereby improving resolution generalization. We further introduce a simple random resize-and-crop augmentation to strengthen order modeling and add micro-conditioning to indicate the applied cropping pattern. On the ImageNet dataset, RPE-2D achieves state-of-the-art resolution generalization performance, outperforming competitive methods when trained at 256^2 and evaluated at 384^2 and 512^2, and when trained at 512^2 and evaluated at 768^2 and 1024^2. RPE-2D also exhibits outstanding capabilities in low-resolution image generation, multi-stage training acceleration, and multi-resolution inheritance.

PDF Details DOI

EAAI Journal 2025 Journal Article

Model reconstruction and update method for dynamic prediction of loader loading resistance using deep incremental learning

Binyun Wu
Liang Hou
Shaojie Wang
Xiangjian Bu
Cheng Xiang

Addressing the challenges of high computational demand and slow speed in Discrete Element Method-Multibody Dynamics (DEM-MBD) co-simulation models, this paper proposes a model reconstruction and update method using deep incremental learning. First, a digital model based on DEM-MBD co-simulation is developed through calibration and verification of experimental data from loader loading operations. Subsequently, feature parameters for DEM-MBD model reconstruction are obtained via analysis of loading resistance-related parameters. Second, an improved Transformer-based deep learning model is trained using simulation data, incorporating knowledge distillation (KD) and domain adaptation transfer (DA) to construct the deep incremental learning model Transformer-KD-DA, aimed at achieving reconstruction and update of the DEM-MBD co-simulation model. Finally, the model performance is verified using simulation data from various loading trajectories. The results demonstrate that the Transformer-KD-DA proposed in this study effectively reconstructs and updates the DEM-MBD co-simulation model, enabling dynamic prediction of loader loading resistance under different loading trajectories.

Details DOI

AAAI Conference 2024 Conference Paper

PDE+: Enhancing Generalization via PDE with Adaptive Distributional Diffusion

Yige Yuan
Bingbing Xu
Bo Lin
Liang Hou
Fei Sun
Huawei Shen
Xueqi Cheng

The generalization of neural networks is a central challenge in machine learning, especially concerning the performance under distributions that differ from training ones. Current methods, mainly based on the data-driven paradigm such as data augmentation, adversarial training, and noise injection, may encounter limited generalization due to model non-smoothness. In this paper, we propose to investigate generalization from a Partial Differential Equation (PDE) perspective, aiming to enhance it directly through the underlying function of neural networks, rather than focusing on adjusting input data. Specifically, we first establish the connection between neural network generalization and the smoothness of the solution to a specific PDE, namely transport equation. Building upon this, we propose a general framework that introduces adaptive distributional diffusion into transport equation to enhance the smoothness of its solution, thereby improving generalization. In the context of neural networks, we put this theoretical framework into practice as PDE+ (PDE with Adaptive Distributional Diffusion) which diffuses each sample into a distribution covering semantically similar inputs. This enables better coverage of potentially unobserved distributions in training, thus improving generalization beyond merely data-driven methods. The effectiveness of PDE+ is validated through extensive experimental settings, demonstrating its superior performance compared to state-of-the-art methods. Our code is available at https://github.com/yuanyige/pde-add.

PDF Details DOI

EAAI Journal 2024 Journal Article

Research on predictive modeling method of loader working resistance in a sensor-less environment

Shaojie Wang
Shuilin Huang
Liang Hou
Tianlin Hu
Jifang Li
Jian Liu

In view of the inconvenient installation and high cost of the current multi-sensor data prediction methods for predicting loader working resistance, this study proposes a method oriented towards predicting loader working resistance in environments with fewer sensors. First, building on previous research (Wu et al. , 2023), non-essential sensor features are removed by a maximum information coefficient (MIC)method that incorporates expert experience. Second, the Optuna automation framework is embedded to realize the training and testing of the proposed method and compare its prediction performance with other popular methods. Finally, in order to verify its generalization performance, it is validated using loader operation data under different working conditions. The results of this study demonstrate that the proposed method effectively and accurately characterizes the work resistance of loaders under operating conditions. With short testing times and excellent generalization performance, the method proves highly applicable and valuable.

Details DOI

NeurIPS Conference 2023 Conference Paper

Augmentation-Aware Self-Supervision for Data-Efficient GAN Training

Liang Hou
Qi Cao
Yige Yuan
Songtao Zhao
Chongyang Ma
Siyuan Pan
Pengfei Wan
Zhongyuan Wang

Training generative adversarial networks (GANs) with limited data is challenging because the discriminator is prone to overfitting. Previously proposed differentiable augmentation demonstrates improved data efficiency of training GANs. However, the augmentation implicitly introduces undesired invariance to augmentation for the discriminator since it ignores the change of semantics in the label space caused by data transformation, which may limit the representation learning ability of the discriminator and ultimately affect the generative modeling performance of the generator. To mitigate the negative impact of invariance while inheriting the benefits of data augmentation, we propose a novel augmentation-aware self-supervised discriminator that predicts the augmentation parameter of the augmented data. Particularly, the prediction targets of real data and generated data are required to be distinguished since they are different during training. We further encourage the generator to adversarially learn from the self-supervised discriminator by generating augmentation-predictable real and not fake data. This formulation connects the learning objective of the generator and the arithmetic $-$ harmonic mean divergence under certain assumptions. We compare our method with state-of-the-art (SOTA) methods using the class-conditional BigGAN and unconditional StyleGAN2 architectures on data-limited CIFAR-10, CIFAR-100, FFHQ, LSUN-Cat, and five low-shot datasets. Experimental results demonstrate significant improvements of our method over SOTA methods in training data-efficient GANs.

PDF Details

ICML Conference 2022 Conference Paper

Conditional GANs with Auxiliary Discriminative Classifier

Liang Hou
Qi Cao 0005
Huawei Shen
Siyuan Pan
Xiaoshuang Li
Xueqi Cheng

Conditional generative models aim to learn the underlying joint distribution of data and labels to achieve conditional data generation. Among them, the auxiliary classifier generative adversarial network (AC-GAN) has been widely used, but suffers from the problem of low intra-class diversity of the generated samples. The fundamental reason pointed out in this paper is that the classifier of AC-GAN is generator-agnostic, which therefore cannot provide informative guidance for the generator to approach the joint distribution, resulting in a minimization of the conditional entropy that decreases the intra-class diversity. Motivated by this understanding, we propose a novel conditional GAN with an auxiliary discriminative classifier (ADC-GAN) to resolve the above problem. Specifically, the proposed auxiliary discriminative classifier becomes generator-aware by recognizing the class-labels of the real data and the generated data discriminatively. Our theoretical analysis reveals that the generator can faithfully learn the joint distribution even without the original discriminator, making the proposed ADC-GAN robust to the value of the coefficient hyperparameter and the selection of the GAN loss, and stable during training. Extensive experimental results on synthetic and real-world datasets demonstrate the superiority of ADC-GAN in conditional generative modeling compared to state-of-the-art classifier-based and projection-based conditional GANs.

Details

AAAI Conference 2021 Conference Paper

SDGNN: Learning Node Representation for Signed Directed Networks

Junjie Huang
Huawei Shen
Liang Hou
Xueqi Cheng

Network embedding is aimed at mapping nodes in a network into low-dimensional vector representations. Graph Neural Networks (GNNs) have received widespread attention and lead to state-of-the-art performance in learning node representations. However, most GNNs only work in unsigned networks, where only positive links exist. It is not trivial to transfer these models to signed directed networks, which are widely observed in the real world yet less studied. In this paper, we first review two fundamental sociological theories (i. e. , status theory and balance theory) and conduct empirical studies on real-world datasets to analyze the social mechanism in signed directed networks. Guided by related sociological theories, we propose a novel Signed Directed Graph Neural Networks model named SDGNN to learn node embeddings for signed directed networks. The proposed model simultaneously reconstructs link signs, link directions, and signed directed triangles. We validate our model’s effectiveness on five real-world datasets, which are commonly used as the benchmark for signed network embeddings. Experiments demonstrate the proposed model outperforms existing models, including feature-based methods, network embedding methods, and several GNN methods.

PDF Details

NeurIPS Conference 2021 Conference Paper

Self-Supervised GANs with Label Augmentation

Liang Hou
Huawei Shen
Qi Cao
Xueqi Cheng

Recently, transformation-based self-supervised learning has been applied to generative adversarial networks (GANs) to mitigate catastrophic forgetting in the discriminator by introducing a stationary learning environment. However, the separate self-supervised tasks in existing self-supervised GANs cause a goal inconsistent with generative modeling due to the fact that their self-supervised classifiers are agnostic to the generator distribution. To address this problem, we propose a novel self-supervised GAN that unifies the GAN task with the self-supervised task by augmenting the GAN labels (real or fake) via self-supervision of data transformation. Specifically, the original discriminator and self-supervised classifier are unified into a label-augmented discriminator that predicts the augmented labels to be aware of both the generator distribution and the data distribution under every transformation, and then provide the discrepancy between them to optimize the generator. Theoretically, we prove that the optimal generator could converge to replicate the real data distribution. Empirically, we show that the proposed method significantly outperforms previous self-supervised and data augmentation GANs on both generative modeling and representation learning across benchmark datasets.

PDF Details

AAAI Conference 2021 Conference Paper

Slimmable Generative Adversarial Networks

Liang Hou
Zehuan Yuan
Lei Huang
Huawei Shen
Xueqi Cheng
Changhu Wang

Generative adversarial networks (GANs) have achieved remarkable progress in recent years, but the continuously growing scale of models makes them challenging to deploy widely in practical applications. In particular, for real-time generation tasks, different devices require generators of different sizes due to varying computing power. In this paper, we introduce slimmable GANs (SlimGANs), which can flexibly switch the width of the generator to accommodate various quality-efficiency trade-offs at runtime. Specifically, we leverage multiple discriminators that share partial parameters to train the slimmable generator. To facilitate the consistency between generators of different widths, we present a stepwise inplace distillation technique that encourages narrow generators to learn from wide ones. As for class-conditional generation, we propose a sliceable conditional batch normalization that incorporates the label information into different widths. Our methods are validated, both quantitatively and qualitatively, by extensive experiments and a detailed ablation study.

PDF Details

ECAI Conference 2020 Conference Paper

Dual Rejection Sampling for Wasserstein Auto-Encoders

Liang Hou
Huawei Shen
Xueqi Cheng

Deep generative models enhanced by Wasserstein distance have achieved remarkable success in recent years. Wasserstein Auto-Encoders (WAEs) are auto-encoder based generative models that aim to minimize the Wasserstein distance between the data distribution and the generated distribution. The quality of generated samples of WAE depends on the distance between the data distribution and the generated distribution. However, WAE actually minimizes a Wasserstein distance between the data distribution and the reconstructed distribution in data space plus a penalty divergence between the aggregated posterior and the prior in latent space, leading a gap between theory and practice. Consequently, the quality of generated samples of WAE is not satisfactory. In this paper, we propose a novel dual rejection sampling method to improve the performance of WAE on the generated samples in the sampling phase. The proposed method first corrects the generative prior by a discriminator based rejection sampling scheme in latent space and then rectifies the generated distribution by another discriminator based rejection sampling method in data space. Our method is validated, both qualitatively and quantitatively, by extensive experiments on three real-world datasets.

Details