Author name cluster

Kai Ye

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.

11 papers

2 author rows

AAAI Conference 2026 Conference Paper

RIS-LAD: A Benchmark and Model for Referring Image Segmentation in Low-Altitude Drone Imagery

Kai Ye
YingShi Luan
Zhudi Chen
Guangyue Meng
Pingyang Dai
Liujuan Cao

Referring Image Segmentation (RIS), which aims to segment specific objects based on natural language descriptions, plays an essential role in vision-language understanding. Despite its progress in remote sensing applications, RIS under Low-Altitude Drone (LAD) scenarios remains underexplored, as existing datasets and methods are typically designed for high-altitude and static-view imagery. They struggled to handle the unique characteristics of LAD views, such as diverse viewpoints and high object density. In this paper, we propose RIS-LAD, the first fine-grained RIS benchmark tailored for LAD scenarios, featuring 13,871 meticulously annotated image-text-mask triplets collected from real-world drone footage with emphasis on small, densely cluttered objects and multi-view perspectives. Additionally, we propose the Semantic-Aware Adaptive Reasoning Network, which decomposes and adaptively routes semantic information to different network stages rather than uniformly injecting all linguistic features. Specifically, the Category-Dominated Linguistic Enhancement aligns visual features with object categories during early encoding, while the Adaptive Reasoning Fusion Module dynamically selects semantic cues across scales to enhance reasoning in complex scenes. Extensive experiments reveal that RIS-LAD presents substantial challenges to state-of-the-art RIS algorithms, and also demonstrate the effectiveness of our proposed model in addressing these challenges.

PDF Details DOI

AAAI Conference 2026 Conference Paper

S²Teacher: Step-by-step Teacher for Sparsely Annotated Oriented Object Detection

Yu Lin
Jianghang Lin
Kai Ye
You Shen
Shengchuan Zhang
Liujuan Cao

Although fully-supervised oriented object detection has made significant progress in remote sensing image understanding, it comes at the cost of labor-intensive annotation. Recent studies have explored weakly and semi-supervised learning to alleviate this burden. However, these methods overlook the difficulties posed by dense annotations in complex remote sensing scenes. In this paper, we introduce a novel setting called sparsely annotated oriented object detection (SAOOD), which only labels partial instances, and propose a solution to address its challenges. Specifically, we focus on two key issues in the setting: (1) sparse labeling leading to overfitting on limited foreground representations, and (2) unlabeled objects (false negatives) confusing feature learning. To this end, we propose the S2Teacher, a novel angle-consistency guided method that progressively mines pseudo-labels for unlabeled objects from easy to hard, enhancing foreground representations. Additionally, it reweights the loss of unlabeled objects to mitigate their impact during training. Extensive experiments demonstrate that S2Teacher not only significantly improves detector performance across different sparse annotation levels but also achieves near-fully-supervised performance on the DOTA dataset with only 10% annotation instances, effectively balancing accuracy and labeling cost.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

AdaDetectGPT: Adaptive Detection of LLM-Generated Text with Statistical Guarantees

Hongyi Zhou
Jin Zhu
Pingfan Su
Kai Ye
Ying Yang
Shakeel Gavioli-Akilagun
Chengchun Shi

We study the problem of determining whether a piece of text has been authored by a human or by a large language model (LLM). Existing state of the art logits-based detectors make use of statistics derived from the log-probability of the observed text evaluated using the distribution function of a given source LLM. However, relying solely on log probabilities can be sub-optimal. In response, we introduce AdaDetectGPT -- a novel classifier that adaptively learns a witness function from training data to enhance the performance of logits-based detectors. We provide statistical guarantees on its true positive rate, false positive rate, true negative rate and false negative rate. Extensive numerical studies show AdaDetectGPT nearly uniformly improves the state-of-the-art method in various combination of datasets and LLMs, and the improvement can reach up to 37\%. A python implementation of our method is available at https: //github. com/Mamba413/AdaDetectGPT.

PDF Details

AAAI Conference 2025 Conference Paper

CAKE: Category Aware Knowledge Extraction for Open-Vocabulary Object Detection

Shiyuan Ma
Donglin Qian
Kai Ye
Shengchuan Zhang

Open vocabulary object detection (OVOD) task aims to detect objects of novel categories beyond the base categories in the training set. To this end, the detector needs to access image-text pairs containing rich semantic information or the visual language pre-trained model (VLM) learned on them. Recent OVOD methods rely on knowledge distillation from VLMs. However, there are two main problems in current methods: (1) Current knowledge distillation frameworks fail to take advantage of the global category information of VLMs and thus fail to learn category-specific knowledge. (2) Due to the overfitting phenomenon of base categories during training, current OVOD networks generally have the problem of suppressing novel categories as background. To address these two problems, we propose a Category Aware Knowledge Extraction framework (CAKE), which consists of a Category-Specific Knowledge Distillation branch (CSKD) and a Category Generalization Region Proposal Network (CG-RPN). CSKD can more fully extract category-strong related information through category-specific distillation, and it is also conducive to filtering the exclusion problem between individuals of the same category; in this process, the model constructs a category-specific feature set to maintain high-quality category features. CG-RPN leverages the guidance of feature set to adjust the confidence scores of region proposals, thereby mining proposals that potentially contain novel categories of objects. Extensive experiments show that our method can plug and play well with many existing methods and significantly improve their detection performance. Moreover, our CAKE framework can reach the-state-of-the-art performance on OV-COCO and OV-LVIS datasets.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Doubly Robust Alignment for Large Language Models

Erhan Xu
Kai Ye
Hongyi Zhou
Luhan Zhu
Francesco Quinzan
Chengchun Shi

This paper studies reinforcement learning from human feedback (RLHF) for aligning large language models with human preferences. While RLHF has demonstrated promising results, many algorithms are highly sensitive to misspecifications in the underlying preference model (e. g. , the Bradley-Terry model), the reference policy, or the reward function, resulting in undesirable fine-tuning. To address model misspecification, we propose a doubly robust preference optimization algorithm that remains consistent when either the preference model or the reference policy is correctly specified (without requiring both). Our proposal demonstrates superior and more robust performance than state-of-the-art algorithms, both in theory and in practice. The code is available at https: //github. com/DRPO4LLM/DRPO4LLM

PDF Details

NeurIPS Conference 2025 Conference Paper

E2E-VGuard: Adversarial Prevention for Production LLM-based End-To-End Speech Synthesis

Zhisheng Zhang
Derui Wang
Yifan Mi
Zhiyong Wu
Jie Gao
Yuxin Cao
Kai Ye
Minhui Xue

Recent advancements in speech synthesis technology have enriched our daily lives, with high-quality and human-like audio widely adopted across real-world applications. However, malicious exploitation like voice-cloning fraud poses severe security risks. Existing defense techniques struggle to address the production large language model (LLM)-based speech synthesis. While previous studies have considered the protection for fine-tuning synthesizers, they assume manually annotated transcripts. Given the labor intensity of manual annotation, end-to-end (E2E) systems leveraging automatic speech recognition (ASR) to generate transcripts are becoming increasingly prevalent, e. g. , voice cloning via commercial APIs. Therefore, this E2E speech synthesis also requires new security mechanisms. To tackle these challenges, we propose E2E-VGuard, a proactive defense framework for two emerging threats: (1) production LLM-based speech synthesis, and (2) the novel attack arising from ASR-driven E2E scenarios. Specifically, we employ the encoder ensemble with a feature extractor to protect timbre, while ASR-targeted adversarial examples disrupt pronunciation. Moreover, we incorporate the psychoacoustic model to ensure perturbative imperceptibility. For a comprehensive evaluation, we test 16 open-source synthesizers and 3 commercial APIs across Chinese and English datasets, confirming E2E-VGuard's effectiveness in timbre and pronunciation protection. Real-world deployment validation is also conducted. Our code and demo page are available at https: //wxzyd123. github. io/e2e-vguard/.

PDF Details

ICLR Conference 2025 Conference Paper

How Far Are We from True Unlearnability?

Kai Ye
Liangcai Su
Chenxiong Qian

High-quality data plays an indispensable role in the era of large models, but the use of unauthorized data for model training greatly damages the interests of data owners. To overcome this threat, several unlearnable methods have been proposed, which generate unlearnable examples (UEs) by compromising the training availability of data. Clearly, due to unknown training purposes and the powerful representation learning capabilities of existing models, these data are expected to be unlearnable for models across multiple tasks, i.e., they will not help improve the model's performance. However, unexpectedly, we find that on the multi-task dataset Taskonomy, UEs still perform well in tasks such as semantic segmentation, failing to exhibit $\textit{cross-task unlearnability}$. This phenomenon leads us to question: $\textit{How far are we from attaining truly unlearnable examples?}$ We attempt to answer this question from the perspective of model optimization. To this end, we observe the difference in the convergence process between clean and poisoned models using a simple model architecture. Subsequently, from the loss landscape we find that only a part of the critical parameter optimization paths show significant differences, implying a close relationship between the loss landscape and unlearnability. Consequently, we employ the loss landscape to explain the underlying reasons for UEs and propose Sharpness-Aware Learnability (SAL) to quantify the unlearnability of parameters based on this explanation. Furthermore, we propose an Unlearnable Distance (UD) to measure the unlearnability of data based on the SAL distribution of parameters in clean and poisoned models. Finally, we conduct benchmark tests on mainstream unlearnable methods using the proposed UD, aiming to promote community awareness of the capability boundaries of existing unlearnable methods.

Details

AAAI Conference 2024 Conference Paper

An Interpretable Approach to the Solutions of High-Dimensional Partial Differential Equations

Lulu Cao
Yufei Liu
Zhenzhong Wang
Dejun Xu
Kai Ye
Kay Chen Tan
Min Jiang

In recent years, machine learning algorithms, especially deep learning, have shown promising prospects in solving Partial Differential Equations (PDEs). However, as the dimension increases, the relationship and interaction between variables become more complex, and existing methods are difficult to provide fast and interpretable solutions for high-dimensional PDEs. To address this issue, we propose a genetic programming symbolic regression algorithm based on transfer learning and automatic differentiation to solve PDEs. This method uses genetic programming to search for a mathematically understandable expression and combines automatic differentiation to determine whether the search result satisfies the PDE and boundary conditions to be solved. To overcome the problem of slow solution speed caused by large search space, we propose a transfer learning mechanism that transfers the structure of one-dimensional PDE analytical solution to the form of high-dimensional PDE solution. We tested three representative types of PDEs, and the results showed that our proposed method can obtain reliable and human-understandable real solutions or algebraic equivalent solutions of PDEs, and the convergence speed is better than the compared methods. Code of this project is at https://github.com/grassdeerdeer/HD-TLGP.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Uncertainty Regularized Evidential Regression

Kai Ye
Tiejin Chen
Hua Wei
Liang Zhan

The Evidential Regression Network (ERN) represents a novel approach that integrates deep learning with Dempster-Shafer's theory to predict a target and quantify the associated uncertainty. Guided by the underlying theory, specific activation functions must be employed to enforce non-negative values, which is a constraint that compromises model performance by limiting its ability to learn from all samples. This paper provides a theoretical analysis of this limitation and introduces an improvement to overcome it. Initially, we define the region where the models can't effectively learn from the samples. Following this, we thoroughly analyze the ERN and investigate this constraint. Leveraging the insights from our analysis, we address the limitation by introducing a novel regularization term that empowers the ERN to learn from the whole training set. Our extensive experiments substantiate our theoretical findings and demonstrate the effectiveness of the proposed solution.

PDF Details DOI

EAAI Journal 2023 Journal Article

Design method for polyurethane-modified asphalt by using Kriging-Particle Swarm Optimization algorithm

Pengzhen Lu
Kai Ye
Tian Jin
Yiheng Ma
Simin Huang
Chenhao Zhou

The preparation process of polyurethane (PU)-modified bitumen involves numerous design parameters and performance response indexes. Due to the variety of polyurethane modifiers, the preparation process of the polyurethane-modified bitumen is not universally applicable. However, the traditional methods such as the response surface method and orthogonal design method have some problems such as low accuracy and a large number of samples required in the preparation process design. Therefore, according to different application environments, the problem of determining the process parameters of the polyurethane-modified bitumen accurately and efficiently needs to be solved urgently. Using Kriging-Particle Swarm Optimization (PSO) algorithm, an efficient process design method for the preparation of polyurethane modified asphalt is proposed in this paper. Combined with the sensitivity analysis method, the relatively sensitive response indexes are screened out to reduce the number of samples and improve the design accuracy. Among them, the dispersion coefficient was evaluated by fluorescence microscopy test using the Christiansen coefficient method to evaluate the uniformity of the dispersed phase of the polyurethane modifier. According to the target performance, the main process parameters of PU modified asphalt were obtained by the Kriging-Particle Swarm Optimization algorithm: shear time 86 min, shear speed 2450 rpm, shear temperature 148 °C, and polyurethane content 18. 6%. The polyurethane-modified bitumen prepared by this optimal process met the expected performance indicators. This study achieved the expected results with a small number of samples, indicating that this method can achieve the purpose of designing the ideal process parameters of polyurethane-modified asphalt efficiently.

Details DOI

NeurIPS Conference 2023 Conference Paper

Learning Visual Prior via Generative Pre-Training

Jinheng Xie
Kai Ye
Yudong Li
Yuexiang Li
Kevin Qinghong Lin
Yefeng Zheng
Linlin Shen
Mike Zheng Shou

Various stuff and things in visual data possess specific traits, which can be learned by deep neural networks and are implicitly represented as the visual prior, e. g. , object location and shape, in the model. Such prior potentially impacts many vision tasks. For example, in conditional image synthesis, spatial conditions failing to adhere to the prior can result in visually inaccurate synthetic results. This work aims to explicitly learn the visual prior and enable the customization of sampling. Inspired by advances in language modeling, we propose to learn Visual prior via Generative Pre-Training, dubbed VisorGPT. By discretizing visual locations, e. g. , bounding boxes, human pose, and instance masks, into sequences, VisorGPT can model visual prior through likelihood maximization. Besides, prompt engineering is investigated to unify various visual locations and enable customized sampling of sequential outputs from the learned prior. Experimental results demonstrate the effectiveness of VisorGPT in modeling visual prior and extrapolating to novel scenes, potentially motivating that discrete visual locations can be integrated into the learning paradigm of current language models to further perceive visual world. Code is available at https: //sierkinhane. github. io/visor-gpt.

PDF Details