Author name cluster

Jian Kang

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.

14 papers

1 author row

AAAI Conference 2026 Conference Paper

DIFFA: Large Language Diffusion Models Can Listen and Understand

Jiaming Zhou
Hongjie Chen
Shiwan Zhao
Jian Kang
Jie Li
Enzhi Wang
Yujie Guo
Haoqin Sun

Recent advances in large language models (LLMs) have shown remarkable capabilities across textual and multimodal domains. In parallel, large language diffusion models have emerged as a promising alternative to the autoregressive paradigm, offering improved controllability, bidirectional context modeling, and robust generation. However, their application to the audio modality remains underexplored. In this work, we introduce DIFFA, the first diffusion-based large audio-language model designed to perform spoken language understanding. DIFFA integrates a frozen diffusion language model with a lightweight dual-adapter architecture that bridges speech understanding and natural language reasoning. We employ a two-stage training pipeline: first, aligning semantic representations via an ASR objective; then, learning instruction-following abilities through synthetic audio-caption pairs automatically generated by prompting LLMs. Despite being trained on only 960 hours of ASR and 127 hours of synthetic instruction data, DIFFA demonstrates competitive performance on major benchmarks, including MMSU, MMAU, and VoiceBench, outperforming several autoregressive open-source baselines. Our results reveal the potential of large language diffusion models for efficient and scalable audio understanding, opening a new direction for speech-driven AI.

PDF Details DOI

AAAI Conference 2026 Conference Paper

WenetSpeech-Yue: A Large-Scale Cantonese Speech Corpus with Multi-dimensional Annotation

Longhao Li
Zhao Guo
Hongjie Chen
Yuhang Dai
Ziyu Zhang
Hongfei Xue
Tianlun Zuo
Chengyou Wang

The development of speech understanding and generation has been significantly accelerated by the availability of large-scale, high-quality speech datasets. Among these, ASR and TTS are regarded as the most established and fundamental tasks. However, for Cantonese (Yue Chinese), spoken by approximately 84.9 million native speakers worldwide, limited annotated resources have hindered progress and resulted in suboptimal ASR and TTS performance. To address this challenge, we propose WenetSpeech-Pipe, an integrated pipeline for building large-scale speech corpus with multi-dimensional annotation tailored for speech understanding and generation. Based on this pipeline, we release WenetSpeech-Yue, the first large-scale Cantonese speech corpus with multi-dimensional annotation for ASR and TTS, covering 21,800 hours across 10 domains with annotations including ASR transcription, text confidence, speaker identity, age, gender, speech quality scores, among other annotations. We also release WSYue-eval, a comprehensive Cantonese benchmark with two components: WSYue-ASR-eval, a manually annotated set for evaluating ASR on short and long utterances, code-switching, and diverse acoustic conditions, and WSYue-TTS-eval, with base and coverage subsets for standard and generalization testing. Experimental results show that models trained on WenetSpeech-Yue achieve competitive results against state-of-the-art (SOTA) Cantonese ASR and TTS systems, including commercial and LLM-based models, highlighting the value of our dataset and pipeline.

PDF Details DOI

JMLR Journal 2025 Journal Article

Bayesian Scalar-on-Image Regression with a Spatially Varying Single-layer Neural Network Prior

Ben Wu
Keru Wu
Jian Kang

Deep neural networks (DNN) have been widely used in scalar-on-image regression to predict an outcome variable from imaging predictors. However, training DNN typically requires large sample sizes for accurate prediction, and the resulting models often lack interpretability. In this work, we propose a novel Bayesian nonlinear scalar-on-image regression framework with a spatially varying single-layer neural network (SV-NN) prior. The SV-NN is constructed using a single hidden layer neural network with its weights generated by the soft-thresholded Gaussian process. Our framework enables the selection of interpretable image regions while achieving high prediction accuracy with limited training samples. The SV-NN offers large prior support for the imaging effect function, facilitating efficient posterior inference on image region selection and automatic network structures determination. We establish the posterior consistency for model parameters and selection consistency for image regions when the number of voxels/pixels grows much faster than the sample size. To ensure computational efficiency, we develop a stochastic gradient Langevin dynamics (SGLD) algorithm for posterior inference. We evaluate our method through extensive comparisons with state-of-the-art deep learning approaches, analyzing multiple real datasets, including task fMRI data from the Adolescent Brain Cognitive Development (ABCD) study. [abs] [ pdf ][ bib ] &copy JMLR 2025. ( edit, beta )

PDF Details

NeurIPS Conference 2025 Conference Paper

CLIMB: Class-imbalanced Learning Benchmark on Tabular Data

Zhining Liu
Zihao Li
Ze Yang
Tianxin Wei
Jian Kang
Yada Zhu
Hendrik Hamann
Jingrui He

Class-imbalanced learning (CIL) on tabular data is important in many real-world applications where the minority class holds the critical but rare outcomes. In this paper, we present CLIMB, a comprehensive benchmark for class-imbalanced learning on tabular data. CLIMB includes 73 real-world datasets across diverse domains and imbalance levels, along with unified implementations of 29 representative CIL algorithms. Built on a high-quality open-source Python package with unified API designs, detailed documentation, and rigorous code quality controls, CLIMB supports easy implementation and comparison between different CIL algorithms. Through extensive experiments, we provide practical insights on method accuracy and efficiency, highlighting the limitations of naive rebalancing, the effectiveness of ensembles, and the importance of data quality. Our code, documentation, and examples are available at https: //github. com/ZhiningLiu1998/imbalanced-ensemble.

PDF Details

TMLR Journal 2025 Journal Article

Fair Online Influence Maximization

Xiangqi Wang
Shaokun Zhang
Jose Efraim Aguilar Escamilla
Qingyun Wu
Xiangliang Zhang
Jian Kang
Huazheng Wang

Fair influence maximization in networks has been actively studied to ensure equity in fields like viral marketing and public health. Existing studies often assume an offline setting, meaning that the learner identifies a set of seed nodes with known per-edge activation probabilities. In this paper, we study the problem of fair online influence maximization, i.e., without knowing the ground-truth activation probabilities. The learner in this problem aims to maximally propagate the information among demographic groups, while interactively selecting seed nodes and observing the activation feedback on the fly. We propose Fair Online Influence Maximization (FOIM) framework that can solve the online influence maximization problem under a wide range of fairness notions. Given a fairness notion, FOIM solves the problem with a combinatorial multi-armed bandit algorithm for balancing exploration-exploitation and an offline fair influence maximization oracle for seed nodes selection. FOIM enjoys sublinear regret when the fairness notion satisfies two mild conditions, i.e., monotonicity and bounded smoothness. Our analyses show that common fairness notions, including maximin fairness, diversity fairness, and welfare function, all satisfy the condition, and we prove the corresponding regret upper bounds under these notions. Extensive empirical evaluations on three real-world networks demonstrate the efficacy of our proposed framework.

PDF Details

NeurIPS Conference 2025 Conference Paper

Understanding and Rectifying Safety Perception Distortion in VLMs

Xiaohan Zou
Jian Kang
George Kesidis
Lu Lin

Recent studies reveal that vision-language models (VLMs) become more susceptible to harmful requests and jailbreak attacks after integrating the vision modality, exhibiting greater vulnerability than their text-only LLM backbones. To uncover the root cause of this phenomenon, we conduct an in-depth analysis and identify a key issue: multimodal inputs introduce an modality-induced activation shift toward a “safer” direction compared to their text-only counterparts, leading VLMs to systematically overestimate the safety of harmful inputs. We refer to this issue as safety perception distortion. To mitigate such distortion, we propose Activation Shift Disentanglement and Calibration (ShiftDC), a training-free method that decomposes and calibrates the modality-induced activation shift to reduce its impact on safety. By isolating and removing the safety-relevant component, ShiftDC restores the inherent safety alignment of the LLM backbone while preserving the vision-language capabilities of VLMs. Experiments demonstrate that ShiftDC significantly enhances safety alignment without impairing model utility.

PDF Details

NeurIPS Conference 2024 Conference Paper

Link prediction is a critical problem in graph learning with broad applications such as recommender systems and knowledge graph completion. Numerous research efforts have been directed at solving this problem, including approaches based on similarity metrics and Graph Neural Networks (GNN). However, most existing solutions are still rooted in conventional supervised learning, which makes it challenging to adapt over time to changing customer interests and to address the inherent dilemma of exploitation versus exploration in link prediction. To tackle these challenges, this paper reformulates link prediction as a sequential decision-making process, where each link prediction interaction occurs sequentially. We propose a novel fusion algorithm, PRB (PageRank Bandits), which is the first to combine contextual bandits with PageRank for collaborative exploitation and exploration. We also introduce a new reward formulation and provide a theoretical performance guarantee for PRB. Finally, we extensively evaluate PRB in both online and offline settings, comparing it with bandit-based and graph-based methods. The empirical success of PRB demonstrates the value of the proposed fusion approach. Our code is released at https: //github. com/jiaruzouu/PRB.

PDF Details DOI

TMLR Journal 2024 Journal Article

Sequential Best-Arm Identification with Application to P300 Speller

Xin Zhou
Botao Hao
Tor Lattimore
Jian Kang
Lexin Li

A brain-computer interface (BCI) is an advanced technology that facilitates direct communication between the human brain and a computer system, by enabling individuals to interact with devices using only their thoughts. The P300 speller is a primary type of BCI system, which allows users to spell words without using a physical keyboard, but instead by capturing and interpreting brain electroencephalogram (EEG) signals under different stimulus presentation paradigms. Traditional non-adaptive presentation paradigms, however, treat each word selection as an isolated event, resulting in a lengthy learning process. To enhance efficiency, we cast the problem as a sequence of best-arm identification tasks within the context of multi-armed bandits, where each task corresponds to the interaction between the user and the system for a single character or word. Leveraging large language models, we utilize the prior knowledge learned from previous tasks to inform and facilitate subsequent tasks. We propose a sequential top-two Thompson sampling algorithm under two scenarios: the fixed-confidence setting and the fixed-budget setting. We study the theoretical property of the proposed algorithm, and demonstrate its substantial empirical improvement through both simulations as well as the data generated from a P300 speller simulator that was built upon the real BCI experiments.

PDF Details

NeurIPS Conference 2024 Conference Paper

Theoretical and Empirical Insights into the Origins of Degree Bias in Graph Neural Networks

Arjun Subramonian
Jian Kang
Yizhou Sun

Graph Neural Networks (GNNs) often perform better for high-degree nodes than low-degree nodes on node classification tasks. This degree bias can reinforce social marginalization by, e. g. , privileging celebrities and other high-degree actors in social networks during social and content recommendation. While researchers have proposed numerous hypotheses for why GNN degree bias occurs, we find via a survey of 38 degree bias papers that these hypotheses are often not rigorously validated, and can even be contradictory. Thus, we provide an analysis of the origins of degree bias in message-passing GNNs with different graph filters. We prove that high-degree test nodes tend to have a lower probability of misclassification regardless of how GNNs are trained. Moreover, we show that degree bias arises from a variety of factors that are associated with a node's degree (e. g. , homophily of neighbors, diversity of neighbors). Furthermore, we show that during training, some GNNs may adjust their loss on low-degree nodes more slowly than on high-degree nodes; however, with sufficiently many epochs of training, message-passing GNNs can achieve their maximum possible training accuracy, which is not significantly limited by their expressive power. Throughout our analysis, we connect our findings to previously-proposed hypotheses for the origins of degree bias, supporting and unifying some while drawing doubt to others. We validate our theoretical findings on 8 common real-world networks, and based on our theoretical and empirical insights, describe a roadmap to alleviate degree bias.

PDF Details DOI

TIST Journal 2023 Journal Article

Highly Efficient Traffic Planning for Autonomous Vehicles to Cross Intersections Without a Stop

Jian Kang
Dan Lin

Waiting in a long queue at traffic lights not only wastes valuable time but also pollutes the environment. With the advances in autonomous vehicles and 5G networks, the previous jamming scenarios at intersections may be turned into non-stop weaving traffic flows. Toward this vision, we propose a highly efficient traffic planning system, namely DASHX, which enables connected autonomous vehicles to cross multi-way intersections without a stop. Specifically, DASHX has a comprehensive model to represent intersections and vehicle status. It can constantly process large volumes of vehicle information, resolve scheduling conflicts, and generate optimal travel plans for all vehicles coming toward the intersection in real time. Unlike existing works that are limited to certain types of intersections and lack considerations of practicability, DASHX is universal for any type of 3D intersection and yields the near-maximum throughput while still ensuring riding comfort. To better evaluate the effectiveness of traffic scheduling systems in real-world scenarios, we developed a sophisticated open source 3D traffic simulation platform (DASHX-SIM) that can handle complicated 3D road layouts and simulate vehicles’ networking and decision-making processes. We have conducted extensive experiments, and the experimental results demonstrate the practicality, effectiveness, and efficiency of the DASHX system and the simulator.

Details DOI

NeurIPS Conference 2023 Conference Paper

Rewrite Caption Semantics: Bridging Semantic Gaps for Language-Supervised Semantic Segmentation

Yun Xing
Jian Kang
Aoran Xiao
Jiahao Nie
Ling Shao
Shijian Lu

Vision-Language Pre-training has demonstrated its remarkable zero-shot recognition ability and potential to learn generalizable visual representations from languagesupervision. Taking a step ahead, language-supervised semantic segmentation enables spatial localization of textual inputs by learning pixel grouping solely from image-text pairs. Nevertheless, the state-of-the-art suffers from a clear semantic gap between visual and textual modalities: plenty of visual concepts appeared in images are missing in their paired captions. Such semantic misalignment circulates in pre-training, leading to inferior zero-shot performance in dense predictions due to insufficient visual concepts captured in textual representations. To close such semantic gap, we propose Concept Curation (CoCu), a pipeline that leverages CLIP to compensate for the missing semantics. For each image-text pair, we establish a concept archive that maintains potential visually-matched concepts with our proposed vision-driven expansion and text-to-vision-guided ranking. Relevant concepts can thus be identified via cluster-guided sampling and fed into pre-training, thereby bridging the gap between visual and textual semantics. Extensive experiments over a broad suite of 8 segmentation benchmarks show that CoCu achieves superb zero-shot transfer performance and greatly boosts language-supervised segmentation baseline by a large margin, suggesting the value of closing semantic gap in pre-training data.

PDF Details

YNIMG Journal 2023 Journal Article

The impact of sociality and affective valence on brain activation: A meta-analysis

Shir Atzil
Ajay B. Satpute
Jiahe Zhang
Michael H. Parrish
Holly Shablack
Jennifer K. MacCormack
Joseph Leshin
Srishti Goel

Thirty years of neuroimaging reveal the set of brain regions consistently associated with pleasant and unpleasant affect in humans-or the neural reference space for valence. Yet some of humans' most potent affective states occur in the context of other humans. Prior work has yet to differentiate how the neural reference space for valence varies as a product of the sociality of affective stimuli. To address this question, we meta-analyzed across 614 social and non-social affective neuroimaging contrasts, summarizing the brain regions that are consistently activated for social and non-social affective information. We demonstrate that across the literature, social and non-social affective stimuli yield overlapping activations within regions associated with visceromotor control, including the amygdala, hypothalamus, anterior cingulate cortex and insula. However, we find that social processing differs from non-social affective processing in that it involves additional cortical activations in the medial prefrontal and posterior cingulum that have been associated with mentalizing and prediction. A Bayesian classifier was able to differentiate unpleasant from pleasant affect, but not social from non-social affective states. Moreover, it was not able to classify unpleasantness from pleasantness at the highest levels of sociality. These findings suggest that highly social scenarios may be equally salient to humans, regardless of their valence.

Details DOI

EAAI Journal 2022 Journal Article

A learning-based model predictive control scheme and its application in biped locomotion

Jingchao Li
Zhaohui Yuan
Sheng Dong
Xiaoyue Sang
Jian Kang

This paper proposes a learning-based model predictive control scheme. This scheme divides the predictive model into a known nominal model and an unknown model residual. Model residual is learned using Gaussian process regression. The learned stochastic model is solved quickly using differential dynamic programming, taking into account control input constraints. The simulation results show that compared with state of art optimal control methods, this scheme has good robustness to model residual, accelerates the solution of high-dimensional problems, and can strictly constrain the control inputs according to the actual situation. Based on this learning-based model predictive control scheme, this paper also proposes an online learning gait generator for the uncertainty problem in the locomotion control of biped robots. The zero moment point is strictly constrained during training to ensure safety. The simulation results show that the gait generator is robust to unknown load and unknown external force.

Details DOI

YNIMG Journal 2016 Journal Article

A depression network of functionally connected regions discovered via multi-attribute canonical correlation graphs

Jian Kang
F. DuBois Bowman
Helen Mayberg
Han Liu

To establish brain network properties associated with major depressive disorder (MDD) using resting-state functional magnetic resonance imaging (Rs-fMRI) data, we develop a multi-attribute graph model to construct a region-level functional connectivity network that uses all voxel level information. For each region pair, we define the strength of the connectivity as the kernel canonical correlation coefficient between voxels in the two regions; and we develop a permutation test to assess the statistical significance. We also construct a network based classifier for making predictions on the risk of MDD. We apply our method to Rs-fMRI data from 20 MDD patients and 20 healthy control subjects in the Predictors of Remission in Depression to Individual and Combined Treatments (PReDICT) study. Using this method, MDD patients can be distinguished from healthy control subjects based on significant differences in the strength of regional connectivity. We also demonstrate the performance of the proposed method using simulationstudies.

Details DOI