AIConfPaper
← Search

Haoyang LI

46 accepted papers

2026

ALIGNING GENERATIVE SPEECH ENHANCEMENT WITH PERCEPTUAL FEEDBACK

Haoyang Li, Nana Hou, Yuchen Hu, Jixun Yao, Sabato Marco Siniscalchi, Norwegian University of Science and Technology +4
ICASSP 2026oral

Language Model (LM)-based speech enhancement (SE) has recently emerged as a promising direction, but existing approaches predominantly rely on token-level likelihood objectives that weakly reflect human perception. This mismatch limits progress, as optimizing signal accuracy does not always improve…

Cited by 0Source ↗PDF ↗Scholar ↗
2026

Adaptive Mixture of Disentangled Experts for Dynamic Graphs under Distribution Shifts

Haibo Chen, Xin Wang, Guanheng Chen, Yuan Meng, Haoyang Li, Yang Yao +5
ICLR 2026poster

Dynamic graph representation learning under distribution shifts has drawn an increasing amount of attention in the research community, given its wide applicability in real-world scenarios. Existing methods typically employ a fixed-architecture design to extract invariant patterns. However, there may…

Cited by 0Source ↗Scholar ↗
2026

Advancing Off-Road Autonomous Driving: The Large-Scale ORAD-3D Dataset and Comprehensive Benchmarks

Chen Min, Jilin Mei, Heng Zhai, Shuai Wang, Tong Sun, Fanjie Kong +14
ICRA 2026poster

A major bottleneck in off-road autonomous driving research lies in the scarcity of large-scale, high-quality datasets and benchmarks. To bridge this gap, we present ORAD-3D, which, to the best of our knowledge, is the largest dataset specifically curated for off-road autonomous driving. ORAD-3D cove…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2026

AudioTrust: Benchmarking The Multifaceted Trustworthiness of Audio Large Language Models

Kai Li, Can Shen, Yile Liu, Jirui Han, Kelong zheng, Xuechao Zou +28
ICLR 2026poster

The rapid development and widespread adoption of Audio Large Language Models (ALLMs) require a rigorous assessment of their trustworthiness. However, existing evaluation frameworks, primarily designed for text, are not equipped to handle the unique vulnerabilities introduced by audio’s acoustic prop…

Cited by 0Source ↗code ↗Scholar ↗
2026

Class-feature Watermark: A Resilient Black-box Watermark Against Model Extraction Attacks

Yaxin Xiao, Qingqing Ye, Zi Liang, Haoyang Li, RongHua Li, Huadi Zheng +1
AAAI 2026technical

Machine learning models constitute valuable intellectual property, yet remain vulnerable to model extraction attacks (MEA), where adversaries replicate their functionality through black-box queries. Model watermarking counters MEAs by embedding forensic markers for ownership verification. Current bl

Cited by 0Source ↗PDF ↗Scholar ↗
2026

Disturbance-Aware Adaptive Compensation in Hybrid Force-Position Locomotion Policy for Legged Robots

Yang Zhang, Zhanxiang Cao, Buqing Nie, Yangqing Fu, Haoyang Li, Zheng Zhang +3
ICRA 2026poster

Reinforcement Learning (RL)-based methods have significantly improved the locomotion performance of legged robots. However, these motion policies face significant challenges when deployed in the real world. Robots operating in uncertain environments struggle to adapt to payload variations and extern…

Cited by 0Source ↗PDF ↗Scholar ↗
2026

From Pixels to Tokens: A Systematic Study of Latent Action Supervision for Vision-Language-Action Models

Yihan Lin, Haoyang Li, Yang Li, Haitao Shen, Yihan Zhao, Chao Shao +1
ICML 2026oral

Latent actions serve as an intermediate representation that enables consistent modeling of vision-language-action (VLA) models across heterogeneous datasets. However, approaches to supervising VLAs with latent actions are fragmented and lack a systematic comparison. This work structures the study of…

Cited by 0Source ↗Scholar ↗
2026

Keep On Going: Learning Robust Humanoid Motion Skills via Selective Adversarial Training

Yang Zhang, Zhanxiang Cao, Buqing Nie, Haoyang Li, Zhong Jiangwei, Qiao Sun +3
AAAI 2026technical

Humanoid robots are expected to operate reliably over long horizons while executing versatile whole-body skills. Yet Reinforcement Learning (RL) motion policies typically lose stability under prolonged operation, sensor/actuator noise, and real world disturbances. In this work, we propose a Selectiv

Cited by 0Source ↗PDF ↗Scholar ↗
2026

Learning Motion Skills with Adaptive Assistive Curriculum Force in Humanoid Robots

Zhanxiang Cao, Yang Zhang, Buqing Nie, Huangxuan Lin, Haoyang Li, Yizhi Chen +2
ICRA 2026poster

Learning policies for complex humanoid tasks remains both challenging and compelling. Inspired by how infants and athletes rely on external support—such as parental walkers or coach-applied guidance—to acquire skills like walking, dancing, and performing acrobatic flips, we propose A2CF: Adaptive As…

Cited by 0Source ↗PDF ↗Scholar ↗
2026

MAKP: Multi-Mode Accurate Kicking Policy for Humanoid Robots

Zheng Zhang, Kaiyang Xu, Zhanxiang Cao, Yizhi Chen, Peng Wang, Haoyang Li +5
ICRA 2026poster

Humanoid robot soccer players face fundamental challenges in achieving stable motion execution and ball trajectory control, particularly under balance constraints during single-leg support phases. In this paper, we introduce MAKP (Multi-mode Accurate Kicking Policy), a novel motion generation-based …

Cited by 0Scholar ↗
2026

Position: Reasoning After Perception Means Reasoning Without Vision

Hongcheng Gao, Zihao Huang, Jingyi Tang, Lin Xu, Xinhao Li, Haoyang Li +9
ICML 2026poster

A common belief in multimodal research is that the perceptual weaknesses of vision--language models can be compensated by stronger language reasoning (e.g., chain-of-thought, in-context learning, or external tools). We challenge this assumption. We argue that for a broad class of visual tasks hard t…

Cited by 0Source ↗Scholar ↗
2026

ReCAPA: Hierarchical Predictive Correction to Mitigate Cascading Failures

Xiyin Zeng, Yuyu Sun, Haoyang Li, Shouqiang Liu, Hao Wang
ICLR 2026poster

Vision–Language–Action (VLA) agents follow instructions to perform multi-step tasks in multimodal environments. To support planning and execution in such settings, many approaches typically adopt structured post-hoc or rely on fixed decomposition and rigid alignment to improve success rate. However,…

Cited by 0Source ↗Scholar ↗
2025

A Sample-Level Evaluation and Generative Framework for Model Inversion Attacks

Haoyang Li, Li Bai, Qingqing Ye, Haibo Hu, Yaxin Xiao, Huadi Zheng +1
AAAI 2025technical

Model Inversion (MI) attacks, which reconstruct the training dataset of neural networks, pose significant privacy concerns in machine learning. Recent MI attacks have managed to reconstruct realistic label-level private data, such as the general appearance of a target person from all training images…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

A Selective Learning Method for Temporal Graph Continual Learning

Hanmo Liu, Shimin Di, Haoyang LI, Xun Jian, Yue Wang, Lei Chen
ICML 2025poster

Node classification is a key task in temporal graph learning (TGL). Real-life temporal graphs often introduce new node classes over time, but existing TGL methods assume a fixed set of classes. This assumption brings limitations, as updating models with full data is costly, while focusing only on ne…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

Anticipate Before Act: Prediction Based Constrained Reinforcement Learning Framework for Skiing Robot Control

Haoyang Li, Xuekang Yang, Jialing Zhu, Zhanxiang Cao, Yang Zhang, Yue Gao
RA-L 2025

Enabling a robot to ski with agility presents an exciting yet complex challenge, primarily due to the intricate dynamics arising from ski-snow interactions. Existing robotic simulators are unable to accurately model the non-rigid, highly dynamic contact between skis and deformable snow surfaces. Hen

Cited by 0Source ↗Scholar ↗
2025

AutoGFM: Automated Graph Foundation Model with Adaptive Architecture Customization

Haibo Chen, Xin Wang, Zeyang Zhang, Haoyang Li, Ling Feng, Wenwu Zhu
ICML 2025oral

Graph foundation models (GFMs) aim to share graph knowledge across diverse domains and tasks to boost graph machine learning. However, existing GFMs rely on hand-designed and fixed graph neural network (GNN) architectures, failing to utilize optimal architectures *w.r.t.* specific domains and tasks…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

DPC: Dual-Prompt Collaboration for Tuning Vision-Language Models

Haoyang Li, Liang Wang, Chao Wang, Jing Jiang, Yan Peng, Guodong Long
CVPR 2025poster

The Base-New Trade-off (BNT) problem universally exists during the optimization of CLIP-based prompt tuning, where continuous fine-tuning on base (target) classes leads to a simultaneous decrease of generalization ability on new (unseen) classes. Existing approaches attempt to regulate the prompt tu…

Cited by 1Source ↗PDF ↗code ↗Scholar ↗
2025

Diffusion Dynamics Models with Generative State Estimation for Cloth Manipulation

Tongxuan Tian, Haoyang Li, Bo Ai, Xiaodi Yuan, Zhiao Huang, Hao Su
CoRL 2025poster

Cloth manipulation is challenging due to its highly complex dynamics, near-infinite degrees of freedom, and frequent self-occlusions, which complicate both state estimation and dynamics modeling. Inspired by recent advances in generative models, we hypothesize that these expressive models can effect…

Cited by 0Source ↗Scholar ↗
2025

Disentangling Invariant Subgraph via Variance Contrastive Estimation under Distribution Shifts

Haoyang Li, Xin Wang, Xueling Zhu, Weigao Wen, Wenwu Zhu
ICML 2025poster

Graph neural networks (GNNs) have achieved remarkable success, yet most are developed under the in-distribution assumption and fail to generalize to out-of-distribution (OOD) environments. To tackle this problem, some graph invariant learning methods aim to learn invariant subgraph against distribut…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

Exposing Numeracy Gaps: A Benchmark to Evaluate Fundamental Numerical Abilities in Large Language Models

Haoyang Li, Xuejia Chen, Zhanchao Xu, Darian Li, Nicole Hu, Fei Teng +5
ACL 2025finding

Large Language Models (LLMs) have demonstrated impressive capabilities in natural language processing tasks, such as text generation and semantic understanding. However, their performance on numerical reasoning tasks, such as basic arithmetic, numerical retrieval, and magnitude comparison, remains s…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

FacLens: Transferable Probe for Foreseeing Non-Factuality in Fact-Seeking Question Answering of Large Language Models

Yanling Wang, Haoyang Li, Hao Zou, Jing Zhang, Xinlei He, Qi Li +1
EMNLP 2025

Despite advancements in large language models (LLMs), non-factual responses still persist in fact-seeking question answering. Unlike extensive studies on post-hoc detection of these responses, this work studies non-factuality prediction (NFP), predicting whether an LLM will generate a non-factual re

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

LLMs Caught in the Crossfire: Malware Requests and Jailbreak Challenges

Haoyang Li, Huan Gao, Zhiyuan Zhao, Zhiyu Lin, Junyu Gao, Xuelong Li
ACL 2025long

The widespread adoption of Large Language Models (LLMs) has heightened concerns about their security, particularly their vulnerability to jailbreak attacks that leverage crafted prompts to generate malicious outputs. While prior research has been conducted on general security capabilities of LLMs, t…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

LlamaPartialSpoof: An LLM-Driven Fake Speech Dataset Simulating Disinformation Generation

Hieu-Thi Luong, Haoyang Li, Lin Zhang, Kong Aik Lee, Eng Siong Chng
ICASSP 2025accepted

Previous fake speech datasets were constructed from a defender’s perspective to develop countermeasure (CM) systems without considering diverse motivations of attackers. To better align with real-life scenarios, we created LlamaPartialSpoof, a 130-hour dataset that contains both fully and partially…

Cited by 26Source ↗Scholar ↗
2025

Reminiscence Attack on Residuals: Exploiting Approximate Machine Unlearning for Privacy

Yaxin Xiao, Qingqing Ye, Li Hu, Huadi Zheng, Haibo Hu, Zi Liang +2
ICCV 2025poster

Machine unlearning enables the removal of specific data from ML models to uphold the *right to be forgotten*. While approximate unlearning algorithms offer efficient alternatives to full retraining, this work reveals that they fail to adequately protect the privacy of unlearned data. In particular,…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

See-Touch-Predict: Active Exploration and Online Perception of Terrain Physics With Legged Robots

Huangxuan Lin, Haoyang Li, Yue Gao
RA-L 2025

Assessing physical properties of the environment with vision helps humans to respond appropriately before entering risky areas. However, equipping robots with such perceptual ability is challenging due to the lack of labeled data. To overcome this challenge, we present the Active Exploration and Onl

Cited by 1Source ↗Scholar ↗
2025

Self-supervised Masked Graph Autoencoder via Structure-aware Curriculum

Haoyang Li, Xin Wang, Zeyang Zhang, Zongyuan Wu, Linxin Xiao, Wenwu Zhu
ICML 2025spotlight

Self-supervised learning (SSL) on graph-structured data has attracted considerable attention recently. Masked graph autoencoder, as one promising generative graph SSL approach that aims to recover masked parts of the input graph data, has shown great success in various downstream graph tasks. Howeve…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

Speech Enhancement Using Continuous Embeddings of Neural Audio Codec

Haoyang Li, Jia Qi Yip, Tianyu Fan, Eng Siong Chng
ICASSP 2025accepted

Recent advancements in Neural Audio Codec (NAC) models have inspired their use in various speech processing tasks, including speech enhancement (SE). In this work, we propose a novel, efficient SE approach by leveraging the pre-quantization output of a pretrained NAC encoder. Unlike prior NAC-based…

Cited by 0Source ↗Scholar ↗
2025

Subgraph Aggregation for Out-of-Distribution Generalization on Graphs

Bowen Liu, Haoyang Li, Shuning Wang, Shuo Nie, Shanghang Zhang
AAAI 2025technical

Out-of-distribution (OOD) generalization in Graph Neural Networks (GNNs) has gained significant attention due to its critical importance in graph-based predictions in real-world scenarios. Existing methods primarily focus on extracting a single causal subgraph from the input graph to achieve general…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQL

Hanbing Liu, Haoyang Li, Xiaokang Zhang, Ruotong Chen, Haiyong Xu, Tian Tian +2
ACL 2025long

Direct Preference Optimization (DPO) has proven effective in complex reasoning tasks like math word problems and code generation. However, when applied to Text-to-SQL datasets, it often fails to improve performance and can even degrade it. Our investigation reveals the root cause: unlike math and co…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

VLR-Driver: Large Vision-Language-Reasoning Models for Embodied Autonomous Driving

Fanjie Kong, Yitong Li, Weihuang Chen, Chen Min, Yizhe Li, Zhiqiang Gao +3
ICCV 2025poster

The rise of embodied intelligence and multi-modal large language models has led to exciting advancements in the field of autonomous driving, establishing it as a prominent research focus in both academia and industry. However, when confronted with intricate and ambiguous traffic scenarios, the lack…

Cited by 0Source ↗PDF ↗Scholar ↗
2024

Accelerating Text-to-Image Editing via Cache-Enabled Sparse Diffusion Inference

Zihao Yu, Haoyang Li, Fangcheng Fu, Xupeng Miao, Bin Cui
AAAI 2024technical

Due to the recent success of diffusion models, text-to-image generation is becoming increasingly popular and achieves a wide range of applications. Among them, text-to-image editing, or continuous text-to-image generation, attracts lots of attention and can potentially improve the quality of generat…

Cited by 13Source ↗PDF ↗code ↗Scholar ↗
2024

CurBench: Curriculum Learning Benchmark

Yuwei Zhou, Zirui Pan, Xin Wang, Hong Chen, Haoyang Li, Yanwen Huang +5
ICML 2024poster

Curriculum learning is a training paradigm where machine learning models are trained in a meaningful order, inspired by the way humans learn curricula. Due to its capability to improve model generalization and convergence, curriculum learning has gained considerable attention and has been widely app…

Cited by 3Source ↗PDF ↗code ↗Scholar ↗
2024

Disentangled Graph Self-supervised Learning for Out-of-Distribution Generalization

Haoyang Li, Xin Wang, Zeyang Zhang, Haibo Chen, Ziwei Zhang, Wenwu Zhu
ICML 2024poster

Graph out-of-distribution (OOD) generalization, aiming to generalize graph neural networks (GNNs) under distribution shifts between training and testing environments, has attracted ever-increasing attention recently. However, existing literature heavily relies on sufficient task-dependent graph labe…

Cited by 11Source ↗PDF ↗Scholar ↗
2024

Graph Invariant Learning with Subgraph Co-mixup for Out-of-Distribution Generalization

Tianrui Jia, Haoyang Li, Cheng Yang, Tao Tao, Chuan Shi
AAAI 2024technical

Graph neural networks (GNNs) have been demonstrated to perform well in graph representation learning, but always lacking in generalization capability when tackling out-of-distribution (OOD) data. Graph invariant learning methods, backed by the invariance principle among defined multiple environments…

Cited by 24Source ↗PDF ↗code ↗Scholar ↗
2024

Multimodal Graph Neural Architecture Search under Distribution Shifts

Jie Cai, Xin Wang, Haoyang Li, Ziwei Zhang, Wenwu Zhu
AAAI 2024technical

Multimodal graph neural architecture search (MGNAS) has shown great success for automatically designing the optimal multimodal graph neural network (MGNN) architecture by leveraging multimodal representation, crossmodal information and graph structure in one unified framework. However, existing MGNA…

Cited by 6Source ↗PDF ↗Scholar ↗
2023

A Generation-based Deductive Method for Math Word Problems

Yuxuan Hu, Jing Zhang, Haoyang Li, Cuiping Li, Hong Chen
EMNLP 2023long main

Math word problems (MWP) involving advanced operators such as linear equation solver cannot be easily tackled by earlier MWP methods, because the existing generation methods suffer from repeated sub-expression generation and deductive methods are restricted to dealing with binary operations. This pa…

Cited by 0Source ↗code ↗Scholar ↗
2023

AutoGT: Automated Graph Transformer Architecture Search

Zizhao Zhang, Xin Wang, Chaoyu Guan, Ziwei Zhang, Haoyang Li, Wenwu Zhu
ICLR 2023top-5%

Although Transformer architectures have been successfully applied to graph data with the advent of Graph Transformer, current design of Graph Transformer still heavily relies on human labor and expertise knowledge to decide proper neural architectures and suitable graph encoding strategies at each T…

Cited by 27Source ↗PDF ↗Scholar ↗
2023

Intent-aware Recommendation via Disentangled Graph Contrastive Learning

Yuling Wang, Xiao Wang, Xiangzhou Huang, Yanhua Yu, Haoyang Li, Mengdi Zhang +2
IJCAI 2023poster

Graph neural network (GNN) based recommender systems have become one of the mainstream trends due to the powerful learning ability from user behavior data. Understanding the user intents from behavior data is the key to recommender systems, which poses two basic requirements for GNN-based recommende…

Cited by 30Source ↗PDF ↗code ↗Scholar ↗
2023

RESDSQL: Decoupling Schema Linking and Skeleton Parsing for Text-to-SQL

Haoyang Li, Jing Zhang, Cuiping Li, Hong Chen
AAAI 2023technical

One of the recent best attempts at Text-to-SQL is the pre-trained language model. Due to the structural property of the SQL queries, the seq2seq model takes the responsibility of parsing both the schema items (i.e., tables and columns) and the skeleton (i.e., SQL keywords). Such coupled targets incr…

Cited by 181Source ↗PDF ↗code ↗Scholar ↗
2023

Spectral Invariant Learning for Dynamic Graphs under Distribution Shifts

Zeyang Zhang, Xin Wang, Ziwei Zhang, Zhou Qin, Weigao Wen, Hui Xue' +2
NeurIPS 2023poster

Dynamic graph neural networks (DyGNNs) currently struggle with handling distribution shifts that are inherent in dynamic graphs. Existing work on DyGNNs with out-of-distribution settings only focuses on the time domain, failing to handle cases involving distribution shifts in the spectral domain. In…

Cited by 16Source ↗PDF ↗code ↗Scholar ↗
2022

Dynamic Graph Neural Networks Under Spatio-Temporal Distribution Shift

Zeyang Zhang, Xin Wang, Ziwei Zhang, Haoyang Li, Zhou Qin, Wenwu Zhu
NeurIPS 2022accept

Dynamic graph neural networks (DyGNNs) have demonstrated powerful predictive abilities by exploiting graph structural and temporal dynamics. However, the existing DyGNNs fail to handle distribution shifts, which naturally exist in dynamic graphs, mainly because the patterns exploited by DyGNNs may b…

Cited by 74Source ↗PDF ↗Scholar ↗
2022

Learning Invariant Graph Representations for Out-of-Distribution Generalization

Haoyang Li, Ziwei Zhang, Xin Wang, Wenwu Zhu
NeurIPS 2022accept

Graph representation learning has shown effectiveness when testing and training graph data come from the same distribution, but most existing approaches fail to generalize under distribution shifts. Invariant learning, backed by the invariance principle from causality, can achieve guaranteed general…

Cited by 134Source ↗PDF ↗Scholar ↗
2021

Disentangled Contrastive Learning on Graphs

Haoyang Li, Xin Wang, Ziwei Zhang, Zehuan Yuan, Hang Li, Wenwu Zhu
NeurIPS 2021poster

Recently, self-supervised learning for graph neural networks (GNNs) has attracted considerable attention because of their notable successes in learning the representation of graph-structure data. However, the formation of a real-world graph typically arises from the highly complex interaction of man…

Cited by 114Source ↗PDF ↗Scholar ↗