Author name cluster

Zeyu Yang

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.

9 papers

2 author rows

AAAI Conference 2026 Conference Paper

SASST: Leveraging Syntax-Aware Chunking and LLMs for Simultaneous Speech Translation

Zeyu Yang
Lai Wei
Roman Koshkin
Xi Chen
Satoshi Nakamura

This work proposes a grammar-based chunking strategy that segments input streams into semantically complete units by parsing dependency relations (e.g., noun phrase boundaries, verb-object structures) and punctuation features. The method ensures chunk coherence and minimizes semantic fragmentation. Building on this mechanism, we present SASST (Syntax-Aware Simultaneous Translation), an end-to-end framework integrating frozen Whisper encoder and decoder-only LLM. The unified architecture dynamically outputs translation tokens or symbols to jointly optimize translation timing and content, with target-side reordering addressing word-order divergence. Experiments on CoVoST2 multilingual corpus (En to De/Zh/Ja) demonstrate significant translation quality improvements across languages, validating the effectiveness of syntactic structures in LLM-driven SimulST systems.

PDF Details DOI

TMLR Journal 2025 Journal Article

Balanced Mixed-Type Tabular Data Synthesis with Diffusion Models

Zeyu Yang
Han Yu
Peikun Guo
Khadija Zanna
Xiaoxue Yang
Akane Sano

Diffusion models have emerged as a robust framework for various generative tasks, including tabular data synthesis. However, current tabular diffusion models tend to inherit bias in the training dataset and generate biased synthetic data, which may influence discriminatory actions. In this research, we introduce a novel tabular diffusion model that incorporates sensitive guidance to generate fair synthetic data with balanced joint distributions of the target label and sensitive attributes, such as sex and race. The empirical results demonstrate that our method effectively mitigates bias in training data while maintaining the quality of the generated samples. Furthermore, we provide evidence that our approach outperforms existing methods for synthesizing tabular data on fairness metrics such as demographic parity ratio and equalized odds ratio, achieving improvements of over $10\%$. Our implementation is available at https://github.com/comp-well-org/fair-tab-diffusion.

PDF Details

ICRA Conference 2025 Conference Paper

LiDARDustX: A LiDAR Dataset for Dusty Unstructured Road Environments

Chenfeng Wei
Qi Wu
Si Zuo
Jiahua Xu
Boyang Zhao
Zeyu Yang
Guotao Xie
Shenhong Wang

Autonomous driving datasets are essential for validating the progress of intelligent vehicle algorithms, which include localization, perception, and prediction. However, existing datasets are predominantly focused on structured urban environments, which limits the exploration of unstructured and specialized scenarios, particularly those characterized by significant dust levels. This paper introduces the LiDARDustX dataset, which is specifically designed for perception tasks under high-dust conditions, such as those encountered in mining areas. The LiDARDustX dataset consists of 30, 000 LiDAR frames captured by six different LiDAR sensors, each accompanied by 3D bounding box annotations and point cloud semantic segmentation. Notably, over 80% of the dataset comprises dust-affected scenes. By utilizing this dataset, we have established a benchmark for evaluating the performance of state-of-the-art 3D detection and segmentation algorithms. Additionally, we have analyzed the impact of dust on perception accuracy and delved into the causes of these effects. The data and further information can be accessed at: https://github.com/vincentweikey/LiDARDustX.

Details

EAAI Journal 2025 Journal Article

Multi-step difference-driven domain adversarial network for few-sample fault detection in dynamic industrial systems

Ruiyi Fang
Kai Wang
Xiaofeng Yuan
Zeyu Yang
Yalin Wang
Chunhua Yang

The escalating production demands for manufacture result in heightened complexity in industrial processes, which leads to frequent changes in operating conditions, thus making few-sample scenarios commonplace. Although many deep learning methods achieve good performance in fault detection tasks, they mostly rely on sufficient data. Therefore, the lack of adequate data presents challenges for accurately representing the process. Moreover, the inherent interplay in processes and among equipment often manifests in data with dynamic characteristic. To address these challenges, we propose a dynamic domain adversarial network (DDAN) for dynamic few-sample fault detection. DDAN based on knowledge transfer, aiming to facilitate modeling data-poor domain with cross-domain information from the data-rich domain. It consists of three main components, a feature extractor, a data reconstructor, and a domain discriminator. To effectively extract features from dynamic samples in few-sample scenarios, a multi-step difference method is introduced. Combined with self-attention, the feature extractor highlights the most significant difference block in the dynamic representations. The output of data reconstructor is utilized for fault detection tasks, while the domain discriminator is applied for domain adaptation with a rebalancing loss. The proposed method is validated on a numerical case and a real-world alumina evaporation process. The experimental results demonstrate an average improvement in the fault detection rate of 2. 9%, with an improvement exceeding 6% for latent variable faults.

Details DOI

ICML Conference 2025 Conference Paper

The Hidden Joules: Evaluating the Energy Consumption of Vision Backbones for Progress Towards More Efficient Model Inference

Zeyu Yang
Wesley Armour

Deep learning has achieved significant success but poses increasing concerns about energy consumption and sustainability. Despite these concerns, there is a lack of understanding of their energy efficiency during inference. In this study, we conduct a comprehensive analysis of the inference energy consumption of 1, 200 ImageNet classification models—the largest evaluation of its kind to date. Our findings reveal a steep decline in accuracy gains relative to the increase in energy usage, highlighting sustainability concerns in the pursuit of marginal improvements. We identify key factors contributing to energy consumption and demonstrate methods to improve energy efficiency. To promote more sustainable AI practices, we introduce an energy efficiency scoring system and develop an interactive web application that allows users to compare models based on accuracy and energy consumption. By providing extensive empirical data and practical tools, we aim to facilitate informed decision-making and encourage collaborative efforts in the development of energy-efficient AI technologies.

Details

EAAI Journal 2024 Journal Article

Additive dynamic Bayesian networks for enhanced feature learning in soft sensor modeling

Junhua Zheng
Lingquan Zeng
Zeyu Yang
Zhiqiang Ge

Due to the advantages of indicating variable structure and efficient reasoning, Bayesian Networks (BN) have been widely used in data-driven soft sensor applications. However, restricted to linear and conditional Gaussian property, BN-based soft sensors rarely achieve high prediction accuracy. In this paper, an ensemble learning framework – additive dynamic Bayesian networks (ADBN) is proposed for enhanced feature learning, in which Dynamic Bayesian networks are used to learn the conditional independencies among variables and construct feature sets for the following base learners. Additional DBNs are constructed upon the residual information from the past model, to carry out feature learning to fit the residuals. The procedure is repeated and a termination rule from the perspective of feature learning is proposed to end this process, and thus the model complexity can be well restricted. The proposed method is validated on two actual industrial cases. It reveals that the ADBN feature learning method has obtained great improvements. Compared to the single DBN feature engineering method, the root mean square error (RMSE) performance has been improved by 20% and 13% on the two cases, respectively.

Details DOI

NeurIPS Conference 2024 Conference Paper

Tetrahedron Splatting for 3D Generation

Chun Gu
Zeyu Yang
Zijie Pan
Xiatian Zhu
Li Zhang

3D representation is essential to the significant advance of 3D generation with 2D diffusion priors. As a flexible representation, NeRF has been first adopted for 3D representation. With density-based volumetric rendering, it however suffers both intensive computational overhead and inaccurate mesh extraction. Using a signed distance field and Marching Tetrahedra, DMTet allows for precise mesh extraction and real-time rendering but is limited in handling large topological changes in meshes, leading to optimization challenges. Alternatively, 3D Gaussian Splatting (3DGS) is favored in both training and rendering efficiency while falling short in mesh extraction. In this work, we introduce a novel 3D representation, Tetrahedron Splatting (TeT-Splatting), that supports easy convergence during optimization, precise mesh extraction, and real-time rendering simultaneously. This is achieved by integrating surface-based volumetric rendering within a structured tetrahedral grid while preserving the desired ability of precise mesh extraction, and a tile-based differentiable tetrahedron rasterizer. Furthermore, we incorporate eikonal and normal consistency regularization terms for the signed distance field to improve generation quality and stability. Critically, our representation can be trained without mesh extraction, making the optimization process easier to converge. Our TeT-Splatting can be readily integrated in existing 3D generation pipelines, along with polygonal mesh for texture optimization. Extensive experiments show that our TeT-Splatting strikes a superior tradeoff among convergence speed, render efficiency, and mesh quality as compared to previous alternatives under varying 3D generation settings.

PDF Details DOI

NeurIPS Conference 2022 Conference Paper

DeepInteraction: 3D Object Detection via Modality Interaction

Zeyu Yang
Jiaqi Chen
Zhenwei Miao
Wei Li
Xiatian Zhu
Li Zhang

Existing top-performance 3D object detectors typically rely on the multi-modal fusion strategy. This design is however fundamentally restricted due to overlooking the modality-specific useful information and finally hampering the model performance. To address this limitation, in this work we introduce a novel modality interaction strategy where individual per-modality representations are learned and maintained throughout for enabling their unique characteristics to be exploited during object detection. To realize this proposed strategy, we design a DeepInteraction architecture characterized by a multi-modal representational interaction encoder and a multi-modal predictive interaction decoder. Experiments on the large-scale nuScenes dataset show that our proposed method surpasses all prior arts often by a large margin. Crucially, our method is ranked at the first position at the highly competitive nuScenes object detection leaderboard.

PDF Details

ICRA Conference 2022 Conference Paper

Instinctive Real-time sEMG-based Control of Prosthetic Hand with Reduced Data Acquisition and Embedded Deep Learning Training

Zeyu Yang
Angus B. Clark
Digby Chappell
Nicolás Rojas 0002

Achieving instinctive multi-grasp control of prosthetic hands typically still requires a large number of sensors, such as electromyography (EMG) electrodes mounted on a residual limb, that can be costly and time consuming to position, with their signals difficult to classify. Deep-learning-based EMG classifiers however have shown promising results over traditional methods, yet due to high computational requirements, limited work has been done with in-prosthetic training. By targeting specific muscles non-invasively, separating grasping action into hold and release states, and implementing data augmentation, we show in this paper that accurate results for embedded, instinctive, multi-grasp control can be achieved with only 2 low-cost sensors, a simple neural network, and minimal amount of training data. The presented controller, which is based on only 2 surface EMG (sEMG) channels, is implemented in an enhanced version of the OLYMPIC prosthetic hand. Results demonstrate that the controller is capable of identifying all 7 specified grasps and gestures with 93% accuracy, and is successful in achieving several real-life tasks in a real world setting.

Details