Author name cluster

Xiyu Wang

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.

4 papers

2 author rows

AAAI Conference 2026 Conference Paper

Rep Deep & Machine Learning: Exemplar-Free Continual Video Action Recognition via Slow-Fast Collaborative Learning

Xueyi Zhang
Chengwei Zhang
Zheng Li
Xiyu Wang
Siqi Cai
Mingrui Lao
Yanming Guo
Huiping Zhuang

In real-world applications, video action recognition models must continuously learn new action categories while retaining previously acquired knowledge. However, most existing approaches rely on storing historical data for replay, which introduces storage burdens and raises data privacy concerns. To address these challenges, we investigate the problem of Exemplar-Free Continual Video Action Recognition (EF-CVAR) and propose a novel framework named Slow-Fast Collaborative Learning (SFCL). SFCL integrates two complementary learning paradigms: a slow branch based on gradient-driven deep learning, which provides strong adaptability to new tasks, and a fast branch based on analytic learning (e.g., Recursive Least Squares), which efficiently preserves old knowledge without requiring access to past samples. To enable effective collaboration between the two branches, we design the Slow-Fast Dynamic Re-parameterization (SFDR) mechanism for adaptive fusion, and the Knowledge Reflection Mechanism (KRM), which mitigates forgetting and task-recency bias via pseudo-feature generation and dual-level knowledge distillation. Extensive experiments on UCF101, HMDB51, and Something-Something V2 demonstrate that SFCL achieves superior performance compared to existing replay-based methods, despite being exemplar-free. Notably, in long-duration continual learning scenarios, SFCL exhibits remarkable robustness, achieving up to a 30.39\% improvement in accuracy over baselines while maintaining a low forgetting rate, highlighting its scalability and effectiveness in real-world video recognition tasks.

PDF Details DOI

IROS Conference 2025 Conference Paper

A Partition-Learning-Selection-Augmentation (PLSA) Framework to Solve Forward Kinematics of Parallel Robots

Ruiqi Xiang
Yongyin Ye
Xiyu Wang
Jindong Xiang
Han Liu
Mengtang Li

The persistent multi-solution challenge in parallel robots’ forward kinematics (FK) has impeded high-precision real-time control. Current data-driven approaches face limitations in predicting accurate and unique solutions, ensuring cross-architectural generalizability, and validating results through continuous trajectory experiments. To address these issues, this work proposes the Partition-Learning-Selection-Augmentation (PLSA) framework, which systematically resolves FK multi-solution challenges. PLSA clusters potential solutions through data partitioning, predicts all feasible solutions in parallel using deep neural networks (DNNs), integrates a selection mechanism to identify optimal solutions, and refines accuracy via the Newton-Raphson method. Cross-configuration tests on Stewart and 3-RRS parallel robots validate PLSA’s adaptability to different architectures, achieving at least 98. 99% accuracy and a computation speed of approximately 30Hz. Additionally, three neural networks (CNN, KAN, and Transformer) are implemented and compared in the Learning-based Selection module, demonstrating PLSA’s generalizability across diverse networks. Comparative studies against analytical, numerical iterative, and prior data-driven methods confirm PLSA’s unique multi-solution resolution capability, delivering submillimeter accuracy with millisecond-level computation, thus establishing a real-time FK calculation methodology.

Details

IJCAI Conference 2024 Conference Paper

Boosting Diffusion Models with an Adaptive Momentum Sampler

Xiyu Wang
Anh-Dung Dinh
Daochang Liu
Chang Xu

Diffusion probabilistic models (DPMs) have been shown to generate high-quality images without the need for delicate adversarial training. The sampling process of DPMs is mathematically similar to Stochastic Gradient Descent (SGD), with both being iteratively updated with a function increment. Building on this, we present a novel reverse sampler for DPMs in this paper, drawing inspiration from the widely-used Adam optimizer. Our proposed sampler can be readily applied to a pre-trained diffusion model, utilizing momentum mechanisms and adaptive updating to enhance the generated image's quality. By effectively reusing update directions from early steps, our proposed sampler achieves a better balance between high-level semantics and low-level details. Additionally, this sampler is flexible and can be easily integrated into pre-trained DPMs regardless of the sampler used during training. Our experimental results on multiple benchmarks demonstrate that our proposed reverse sampler yields remarkable improvements over different baselines.

PDF Details DOI

ICML Conference 2024 Conference Paper

Bridging Data Gaps in Diffusion Models with Adversarial Noise-Based Transfer Learning

Xiyu Wang
Baijiong Lin
Daochang Liu
Ying-Cong Chen
Chang Xu 0002

Diffusion Probabilistic Models (DPMs) show significant potential in image generation, yet their performance hinges on having access to large datasets. Previous works, like Generative Adversarial Networks (GANs), have tackled the limited data problem by transferring pre-trained models learned with sufficient data. However, those methods are hard to be utilized in DPMs since the distinct differences between DPM-based and GAN-based methods, showing in the unique iterative denoising process integral and the need for many timesteps with no-targeted noise in DPMs. In this paper, we propose a novel DPMs-based transfer learning method, ANT, to address the limited data problem. It includes two strategies: similarity-guided training, which boosts transfer with a classifier, and adversarial noise selection which adaptively chooses targeted noise based on the input image. Extensive experiments in the context of few-shot image generation tasks demonstrate that our method is not only efficient but also excels in terms of image quality and diversity when compared to existing GAN-based and DDPM-based methods.

Details