Author name cluster

Yuning Wu

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.

3 papers

1 author row

EAAI Journal 2026 Journal Article

Cross-domain structural damage identification using frequency guided cycle-consistent generative adversarial network

Xiaohang Zhou
Yuxin Liu
Ranting Cui
Yuning Wu

Discrepancies between Finite Element Model (FEM) simulations and actual measurements result in substantial domain mismatches, posing a significant challenge for model-driven structural damage identification. Existing cross-domain damage identification methods commonly suffer from misalignment between domain-crossing features and damage-related features, as well as unstable network training, thereby limiting their effectiveness. To address these issues and achieve both efficient domain adaptation and high-precision damage identification, this study proposes a Frequency-Guided Cycle-Consistent Generative Adversarial Network (FG-CycleGAN), integrated with a Residual Neural Network (ResNet). First, frequency cosine similarity is introduced into the adversarial training process to quantify spectral discrepancies between generated and measured samples, ensuring the preservation of damage-relevant features during the cross-domain transformation. Subsequently, ResNet is employed to extract essential features from the samples generated by FG-CycleGAN and map them to corresponding structural damage states. To validate the approach, a damage identification experiment is conducted on a steel truss model. Comparative analysis reveals that conventional Adversarial Discriminative Domain Adaptation (ADDA) yields a relatively low F1-score of 0. 35, while the basic CycleGAN achieves 0. 92. In contrast, the proposed FG-CycleGAN further improves performance, attaining an F1-score of 0. 99. The results confirm that FG-CycleGAN not only outperforms existing methods in terms of accuracy but also offers a robust framework for cross-domain structural damage identification.

Details DOI

NeurIPS Conference 2025 Conference Paper

WritingBench: A Comprehensive Benchmark for Generative Writing

Yuning Wu
Jiahao Mei
Ming Yan
Chenliang Li
Shaopeng Lai
Yuran Ren
Zijia Wang
Ji Zhang

Recent advancements in large language models (LLMs) have significantly enhanced text generation capabilities, yet evaluating their performance in generative writing remains a challenge. Existing benchmarks primarily focus on generic text generation or limited in writing tasks, failing to capture the diverse requirements of high-quality written contents across various domains. To bridge this gap, we present WritingBench, a comprehensive benchmark designed to evaluate LLMs across 6 core writing domains and 100 subdomains. We further propose a query-dependent evaluation framework that empowers LLMs to dynamically generate instance-specific assessment criteria. This framework is complemented by a fine-tuned critic model for criteria-aware scoring, enabling evaluations in style, format and length. The framework's validity is further demonstrated by its data curation capability, which enables a 7B-parameter model to outperform the performance of GPT-4o in writing. We open-source the benchmark, along with evaluation tools and modular framework components, to advance the development of LLMs in writing.

PDF Details

AAAI Conference 2024 System Paper

AudioGPT: Understanding and Generating Speech, Music, Sound, and Talking Head

Rongjie Huang
Mingze Li
Dongchao Yang
Jiatong Shi
Xuankai Chang
Zhenhui Ye
Yuning Wu
Zhiqing Hong

Large language models (LLMs) have exhibited remarkable capabilities across a variety of domains and tasks, challenging our understanding of learning and cognition. Despite the recent success, current LLMs are not capable of processing complex audio information or conducting spoken conversations (like Siri or Alexa). In this work, we propose a multi-modal AI system named AudioGPT, which complements LLMs (i.e., ChatGPT) with 1) foundation models to process complex audio information and solve numerous understanding and generation tasks; and 2) the input/output interface (ASR, TTS) to support spoken dialogue. With an increasing demand to evaluate multi-modal LLMs of human intention understanding and cooperation with foundation models, we outline the principles and processes and test AudioGPT in terms of consistency, capability, and robustness. Experimental results demonstrate the capabilities of AudioGPT in solving 16 AI tasks with speech, music, sound, and talking head understanding and generation in multi-round dialogues, which empower humans to create rich and diverse audio content with unprecedented ease. Code can be found in https://github.com/AIGC-Audio/AudioGPT

PDF Details DOI