Author name cluster
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.
EAAI Journal 2026 Journal Article
JBHI Journal 2025 Journal Article
Detecting early ischemic lesions (EIL) in computed tomography (CT) images is crucial for reducing diagnostic time and minimizing neuron loss due to oxygen deprivation. This paper introduces DCTP-Net, a dual-branch network for segmenting acute ischemic stroke lesions in CT images, consisting of a segmentation branch and a prompt-aware branch. The segmentation branch uses an encoder-decoder network as the backbone to identify lesions, where the encoder fuses CT image features with prompt features from the prompt-aware branch. To enhance semantic feature extraction and reduce the impact of cerebral structural details, we introduce a cross-collaboration dynamic connection (CCDC) module to link the encoder and decoder. The prompt-aware branch includes a learnable prompt (LP) block to incorporate cerebral prior knowledge, and the prompt-aware encoder (PAE) combines the LP block with multi-level features from the segmentation branch for more precise representation. Additionally, we propose a CLIP-enhance textual prompt (CETP) module that utilizes the CLIP text encoder to generate specialized convolutional parameters for the segmentation head. These parameters are tailored to the unique characteristics of each input image, improving segmentation performance. Qualitative and quantitative studies reveal that DCTP-Net outperforms the current state-of-the-art, IS-Net, with Dice score increases of 3. 9% on AISD and 3. 8% on ISLES2018, demonstrating its superiority in EIL segmentation.
EAAI Journal 2024 Journal Article
NeurIPS Conference 2023 Conference Paper
In this paper, we present EV-Eye, a first-of-its-kind large scale multimodal eye tracking dataset aimed at inspiring research on high-frequency eye/gaze tracking. EV-Eye utilizes an emerging bio-inspired event camera to capture independent pixel-level intensity changes induced by eye movements, achieving sub-microsecond latency. Our dataset was curated over a two-week period and collected from 48 participants encompassing diverse genders and age groups. It comprises over 1. 5 million near-eye grayscale images and 2. 7 billion event samples generated by two DAVIS346 event cameras. Additionally, the dataset contains 675 thousands scene images and 2. 7 million gaze references captured by Tobii Pro Glasses 3 eye tracker for cross-modality validation. Compared with existing event-based high-frequency eye tracking datasets, our dataset is significantly larger in size, and the gaze references involve more natural eye movement patterns, i. e. , fixation, saccade and smooth pursuit. Alongside the event data, we also present a hybrid eye tracking method as benchmark, which leverages both the near-eye grayscale images and event data for robust and high-frequency eye tracking. We show that our method achieves higher accuracy for both pupil and gaze estimation tasks compared to the existing solution.
IJCAI Conference 2022 Conference Paper
Though deep reinforcement learning (DRL) has obtained substantial success, it may encounter catastrophic failures due to the intrinsic uncertainty of both transition and observation. Most of the existing methods for safe reinforcement learning can only handle transition disturbance or observation disturbance since these two kinds of disturbance affect different parts of the agent; besides, the popular worst-case return may lead to overly pessimistic policies. To address these issues, we first theoretically prove that the performance degradation under transition disturbance and observation disturbance depends on a novel metric of Value Function Range (VFR), which corresponds to the gap in the value function between the best state and the worst state. Based on the analysis, we adopt conditional value-at-risk (CVaR) as an assessment of risk and propose a novel reinforcement learning algorithm of CVaR-Proximal-Policy-Optimization (CPPO) which formalizes the risk-sensitive constrained optimization problem by keeping its CVaR under a given threshold. Experimental results show that CPPO achieves a higher cumulative reward and is more robust against both observation and transition disturbances on a series of continuous control tasks in MuJoCo.
IJCAI Conference 2021 Conference Paper
AlphaZero has achieved superhuman performance on various perfect-information games, such as chess, shogi and Go. However, directly applying AlphaZero to imperfect-information games (IIG) is infeasible, due to the fact that traditional MCTS methods cannot handle missing information of other players. Meanwhile, there have been several extensions of MCTS for IIGs, by implicitly or explicitly sampling a state of other players. But, due to the inability to handle private and public information well, the performance of these methods is not satisfactory. In this paper, we extend AlphaZero to multiplayer IIGs by developing a new MCTS method, Action-Prediction MCTS (AP-MCTS). In contrast to traditional MCTS extensions for IIGs, AP-MCTS first builds the search tree based on public information, adopts the policy-value network to generalize between hidden states, and finally predicts other players' actions directly. This design bypasses the inefficiency of sampling and the difficulty of predicting the state of other players. We conduct extensive experiments on the popular 3-player poker game DouDiZhu to evaluate the performance of AP-MCTS combined with the framework AlphaZero. When playing against experienced human players, AP-MCTS achieved a 65. 65\% winning rate, which is almost twice the human's winning rate. When comparing with state-of-the-art DouDiZhu AIs, the Elo rating of AP-MCTS is 50 to 200 higher than them. The ablation study shows that accurate action prediction is the key to AP-MCTS winning.
IJCAI Conference 2021 Conference Paper
It is important yet challenging to perform accurate and interpretable time series forecasting. Though deep learning methods can boost forecasting accuracy, they often sacrifice interpretability. In this paper, we present a new scheme of series saliency to boost both accuracy and interpretability. By extracting series images from sliding windows of the time series, we design series saliency as a mixup strategy with a learnable mask between the series images and their perturbed versions. Series saliency is model agnostic and performs as an adaptive data augmentation method for training deep models. Moreover, by slightly changing the objective, we optimize series saliency to find a mask for interpretable forecasting in both feature and time dimensions. Experimental results on several real datasets demonstrate that series saliency is effective to produce accurate time-series forecasting results as well as generate temporal interpretations.
IJCAI Conference 2020 Conference Paper
Automatic diagnosis based on clinical notes is critical especially in the emergency department, where a fast and professional result is vital in assuring proper and timely treatment. Previous works formalize this task as plain text classification and fail to utilize the medically significant tree structure of International Classification of Diseases (ICD) coding system. Besides, external medical knowledge is rarely used before, and we explore it by extracting relevant materials from Wikipedia or Baidupedia. In this paper, we propose a knowledge-based tree decoding model (K-BTD), and the inference procedure is a top-down decoding process from the root node to leaf nodes. The stepwise inference procedure enables the model to give support for decision at each step, which visualizes the diagnosis procedure and adds to the interpretability of final predictions. Experiments on real-world data from the emergency department of a large-scale hospital indicate that the proposed model outperforms all baselines in both micro-F1 and macro-F1, and reduce the semantic distance dramatically.
IJCAI Conference 2019 Conference Paper
Identifying adversarial examples is beneficial for understanding deep networks and developing robust models. However, existing attacking methods for image object detection have two limitations: weak transferability---the generated adversarial examples often have a low success rate to attack other kinds of detection methods, and high computation cost---they need much time to deal with video data, where many frames need polluting. To address these issues, we present a generative method to obtain adversarial images and videos, thereby significantly reducing the processing time. To enhance transferability, we manipulate the feature maps extracted by a feature network, which usually constitutes the basis of object detectors. Our method is based on the Generative Adversarial Network (GAN) framework, where we combine a high-level class loss and a low-level feature loss to jointly train the adversarial example generator. Experimental results on PASCAL VOC and ImageNet VID datasets show that our method efficiently generates image and video adversarial examples, and more importantly, these adversarial examples have better transferability, therefore being able to simultaneously attack two kinds of representative object detection models: proposal based models like Faster-RCNN and regression based models like SSD.
IJCAI Conference 2017 Conference Paper
We study envy-free cake cutting with strategic agents, where each agent may manipulate his private information in order to receive a better allocation. We focus on piecewise constant utility functions and consider two scenarios: the general setting without any restriction on the allocations and the restricted setting where each agent has to receive a connected piece. We show that no deterministic truthful envy-free mechanism exists in the connected piece scenario, and the same impossibility result for the general setting with some additional mild assumptions on the allocations. Finally, we study a large market model where the economy is replicated and demonstrate that truth-telling converges to a Nash equilibrium.
AAAI Conference 2017 Conference Paper
Finding semantic attributes to describe related concepts is typically a hard problem. The commonly used attributes in most fields are designed by domain experts, which is expensive and time-consuming. In this paper we propose an efficient method to learn human comprehensible attributes with crowdsourcing. We first design an analogical interface to collect relative labels from the crowds. Then we propose a hierarchical Bayesian model, as well as an efficient initialization strategy, to aggregate labels and extract concise attributes. Our experimental results demonstrate promise on discovering diverse and convincing attributes, which significantly improve the performance of the challenging zero-shot learning tasks.
AAAI Conference 2016 Conference Paper
We present a discriminative nonparametric latent feature relational model (LFRM) for link prediction to automatically infer the dimensionality of latent features. Under the generic RegBayes (regularized Bayesian inference) framework, we handily incorporate the prediction loss with probabilistic inference of a Bayesian model; set distinct regularization parameters for different types of links to handle the imbalance issue in real networks; and unify the analysis of both the smooth logistic log-loss and the piecewise linear hinge loss. For the nonconjugate posterior inference, we present a simple Gibbs sampler via data augmentation, without making restricting assumptions as done in variational methods. We further develop an approximate sampler using stochastic gradient Langevin dynamics to handle large networks with hundreds of thousands of entities and millions of links, orders of magnitude larger than what existing LFRM models can process. Extensive studies on various real networks show promising performance.
AAAI Conference 2016 Conference Paper
In a Fisher market game, a market equilibrium is computed in terms of the utility functions and money endowments that agents reported. As a consequence, an individual buyer may misreport his private information to obtain a utility gain. We investigate the extent to which an agent’s utility can be increased by unilateral strategic plays and prove that the percentage of this improvement is at most 2 for markets with weak gross substitute utilities. Equivalently, we show that truthfully reporting is a 0. 5-approximate Nash equilibrium in this game. To identify sufficient conditions for truthfully reporting being close to Nash equilibrium, we conduct a parameterized study on strategic behaviors and further show that the ratio of utility gain decreases linearly as buyer’s initial endowment increases or his maximum share of an item decreases. Finally, we consider collusive behavior of a coalition and prove that the utility gain is bounded by 1/(1 − maximum share of the collusion). Our findings justify the truthful reporting assumption in Fisher markets by a quantitative study on participants incentive, and imply that under large market assumption, the utility gain of a buyer from manipulations diminishes to 0.
AAAI Conference 2016 Conference Paper
Feature extraction is an important task in machine learning. In this paper, we present a simple and efficient method, named max-margin data shifting (MMDS), to process the data before feature extraction. By relying on a large-margin classifier, MMDS is helpful to enhance the discriminative ability of subsequent feature extractors. The kernel trick can be applied to extract nonlinear features from input data. We further analyze in detail the example of principal component analysis (PCA). The empirical results on multiple linear and nonlinear models demonstrate that MMDS can efficiently improve the performance of unsupervised extractors.
YNICL Journal 2016 Journal Article
JMLR Journal 2014 Journal Article
Existing Bayesian models, especially nonparametric Bayesian methods, rely on specially conceived priors to incorporate domain knowledge for discovering improved latent representations. While priors affect posterior distributions through Bayes' rule, imposing posterior regularization is arguably more direct and in some cases more natural and general. In this paper, we present regularized Bayesian inference (RegBayes), a novel computational framework that performs posterior inference with a regularization term on the desired post-data posterior distribution under an information theoretical formulation. RegBayes is more flexible than the procedure that elicits expert knowledge via priors, and it covers both directed Bayesian networks and undirected Markov networks. When the regularization is induced from a linear operator on the posterior distributions, such as the expectation operator, we present a general convex-analysis theorem to characterize the solution of RegBayes. Furthermore, we present two concrete examples of RegBayes, infinite latent support vector machines (iLSVM) and multi-task infinite latent support vector machines (MT-iLSVM), which explore the large- margin idea in combination with a nonparametric Bayesian model for discovering predictive latent features for classification and multi-task learning, respectively. We present efficient inference methods and report empirical studies on several benchmark data sets, which appear to demonstrate the merits inherited from both large-margin learning and Bayesian nonparametrics. Such results contribute to push forward the interface between these two important subfields, which have been largely treated as isolated in the community. [abs] [ pdf ][ bib ] © JMLR 2014. ( edit, beta )
AAAI Conference 2014 Conference Paper
Dropout and other feature noising schemes have shown promising results in controlling over-fitting by artificially corrupting the training data. Though extensive theoretical and empirical studies have been performed for generalized linear models, little work has been done for support vector machines (SVMs), one of the most successful approaches for supervised learning. This paper presents dropout training for linear SVMs. To deal with the intractable expectation of the non-smooth hinge loss under corrupting distributions, we develop an iteratively re-weighted least square (IRLS) algorithm by exploring data augmentation techniques. Our algorithm iteratively minimizes the expectation of a re-weighted least square problem, where the re-weights have closedform solutions. The similar ideas are applied to develop a new IRLS algorithm for the expected logistic loss under corrupting distributions. Our algorithms offer insights on the connection and difference between the hinge loss and logistic loss in dropout training. Empirical results on several real datasets demonstrate the effectiveness of dropout training on significantly boosting the classification accuracy of linear SVMs.
JMLR Journal 2014 Journal Article
Max-margin learning is a powerful approach to building classifiers and structured output predictors. Recent work on max-margin supervised topic models has successfully integrated it with Bayesian topic models to discover discriminative latent semantic structures and make accurate predictions for unseen testing data. However, the resulting learning problems are usually hard to solve because of the non-smoothness of the margin loss. Existing approaches to building max-margin supervised topic models rely on an iterative procedure to solve multiple latent SVM subproblems with additional mean-field assumptions on the desired posterior distributions. This paper presents an alternative approach by defining a new max-margin loss. Namely, we present Gibbs max-margin supervised topic models, a latent variable Gibbs classifier to discover hidden topic representations for various tasks, including classification, regression and multi-task learning. Gibbs max- margin supervised topic models minimize an expected margin loss, which is an upper bound of the existing margin loss derived from an expected prediction rule. By introducing augmented variables and integrating out the Dirichlet variables analytically by conjugacy, we develop simple Gibbs sampling algorithms with no restrictive assumptions and no need to solve SVM subproblems. Furthermore, each step of the âaugment-and-collapse" Gibbs sampling algorithms has an analytical conditional distribution, from which samples can be easily drawn. Experimental results on several medium-sized and large-scale data sets demonstrate significant improvements on time efficiency. The classification performance is also improved over competitors on binary, multi- class and multi-label classification tasks. [abs] [ pdf ][ bib ] © JMLR 2014. ( edit, beta )
IJCAI Conference 2013 Conference Paper
Relational topic models have shown promise on analyzing document network structures and discovering latent topic representations. This paper presents three extensions: 1) unlike the common link likelihood with a diagonal weight matrix that allows the-same-topic interactions only, we generalize it to use a full weight matrix that captures all pairwise topic interactions and is applicable to asymmetric networks; 2) instead of doing standard Bayesian inference, we perform regularized Bayesian inference with a regularization parameter to deal with the imbalanced link structure issue in common real networks; and 3) instead of doing variational approximation with strict mean-field assumptions, we present a collapsed Gibbs sampling algorithm for the generalized relational topic models without making restricting assumptions. Experimental results demonstrate the significance of these extensions on improving the prediction performance, and the time efficiency can be dramatically improved with a simple fast approximation method.
AAAI Conference 2012 Conference Paper
In cooperative games, a key question is to find a division of payoffs to coalition members in a fair manner. Nucleolus is one of such solution concepts that provides a stable solution for the grand coalition. We study the computation of the nucleolus of a number of cooperative games, including fractional matching games and fractional edge cover games on general weighted graphs, as well as vertex cover games and clique games on weighted bipartite graphs. Our results are on the positive side—we give efficient algorithms to compute the nucleolus, as well as the least core, of all of these games.
AAAI Conference 2012 Conference Paper
We consider the classic cake cutting problem where one allocates a divisible cake to n participating agents. Among all valid divisions, fairness and efficiency (a. k. a. social welfare) are the most critical criteria to satisfy and optimize, respectively. We study computational complexity of computing an efficiency optimal division given the conditions that the allocation satisfies proportional fairness and assigns each agent a connected piece. For linear valuation functions, we give a polynomial time approximation scheme to compute an efficiency optimal allocation. On the other hand, we show that the problem is NP-hard to approximate within a factor of Ω 1 √ n for general piecewise constant functions, and is NP-hard to compute for normalized functions.
IJCAI Conference 2011 Conference Paper
The social lending market, with over a billion dollars in loans, is a two-sided matching market where borrowers specify demands and lenders specify total budgets and their desired interest rates from each acceptable borrower. Because different borrowers correspond to different risk-return profiles, lenders have preferences over acceptable borrowers; a borrower prefers lenders in order of the interest rates they offer to her. We investigate the question of what is a computationally feasible, 'good', allocation to clear this market. We design a strongly polynomial time algorithm for computing a Pareto-efficient stable outcome in a two-sided many-to-many matching market within differences, and use this to compute an allocation for the social lending market that satisfies the properties of stability - a standard notion of fairness in two-sided matching markets - and Pareto efficiency; and additionally addresses envy-freeness amongst similar borrowers and risk diversification for lenders.
IJCAI Conference 2011 Conference Paper
An important task in the analysis of multiagent systems is to understand how groups of selfish players can form coalitions, i. e. , work together in teams. In this paper, we study the dynamics of coalition formation under bounded rationality. We consider settings where each team's profit is given by a concave function, and propose three profit-sharing schemes, each of which is based on the concept of marginal utility. The agents are assumed to be myopic, i. e. , they keep changing teams as long as they can increase their payoff by doing so. We study the properties (such as closeness to Nash equilibrium or total profit) of the states that result after a polynomial number of such moves, and prove bounds on the price of anarchy and the price of stability of the corresponding games.
NeurIPS Conference 2011 Conference Paper
Unlike existing nonparametric Bayesian models, which rely solely on specially conceived priors to incorporate domain knowledge for discovering improved latent representations, we study nonparametric Bayesian inference with regularization on the desired posterior distributions. While priors can indirectly affect posterior distributions through Bayes' theorem, imposing posterior regularization is arguably more direct and in some cases can be much easier. We particularly focus on developing infinite latent support vector machines (iLSVM) and multi-task infinite latent support vector machines (MT-iLSVM), which explore the large-margin idea in combination with a nonparametric Bayesian model for discovering predictive latent features for classification and multi-task learning, respectively. We present efficient inference methods and report empirical studies on several benchmark datasets. Our results appear to demonstrate the merits inherited from both large-margin learning and Bayesian nonparametrics.
NeurIPS Conference 2010 Conference Paper
Learning from multi-view data is important in many applications, such as image classification and annotation. In this paper, we present a large-margin learning framework to discover a predictive latent subspace representation shared by multiple views. Our approach is based on an undirected latent space Markov network that fulfills a weak conditional independence assumption that multi-view observations and response variables are independent given a set of latent variables. We provide efficient inference and parameter estimation methods for the latent subspace model. Finally, we demonstrate the advantages of large-margin learning on real video and web image data for discovering predictive latent representations and improving the performance on image classification, annotation and retrieval.
ICAART Conference 2009 Conference Paper
Learning vector quantization (LVQ) is a supervised learning algorithm for data classification. Since LVQ is based on prototype vectors, it is a neural network approach particularly applicable in non-linear separation problems. Existing LVQ algorithms are mostly focused on numerical data. This paper presents a batch type LVQ algorithm used for mixed numerical and categorical data. Experiments on various data sets demonstrate the proposed algorithm is effective to improve the capability of standard LVQ to deal with categorical data.
SODA Conference 2008 Conference Paper
ICRA Conference 1997 Conference Paper
An experimental study on touch-driven robot control, or tactile-servo, is presented in this paper. The tactile servo scheme uses a tactile feature Jacobian matrix to relate the differential change in tactile feature space to that in the robot task space. Tactile Jacobian matrices are constructed through a finite element (FE) model of the tactile sensor. Detailed implementation of the control scheme and experimental results are presented.
ICRA Conference 1996 Conference Paper
Studies on contact problems such as shape recovery from tactile sensing involve complex geometrics and kinematics. In the first part of the paper we introduce a convenient matrix expression of 3-D surfaces, which can be easily applied to manipulate contact constraints and kinematics equations using homogeneous transformation. In the second part of the paper we apply the notation introduced in the first part and derive an active tactile sensing strategy which consists of a shape sensing algorithm and a touch-based robot control scheme.
IROS Conference 1995 Conference Paper
A novel method to perform edge tracking using tactile sensors is presented in this paper. Using the tactile servo scheme, a robot manipulator is driven only by real-time tactile feedback from the array tactile sensors mounted directly on the robot end-effector. Compared with previous approaches, the control scheme presented in this paper is consistent and more efficient. Real-time edge tracking experiments are conducted using an experimental system consisting of a PUMA 260, a single rigid finger and a planar array tactile sensor. Experimental results show satisfactory control speed and accuracy for both straight and curved edge tracking. An example of active tactile sensing of a unknown object using edge tracking is also demonstrated.
IROS Conference 1995 Conference Paper
A real-time tactile image processing algorithm for edge contact is presented in this paper. Based on basic elasticity results, closed-form solutions for calculating contact force and local contact geometries (i. e. , location and orientation of the line of contact) from the first three moments of a tactile image are derived. Computational complexity of the proposed algorithm and those of the previous approaches are compared, and passive tactile sensing experiments are performed. It is shown that the proposed algorithm has the advantage of persevering force information and is more consistent and computationally efficient.
IROS Conference 1992 Conference Paper
ICRA Conference 1985 Conference Paper
Recent research on fast exact maneuvering strategies for manipulators has employed acceleration commands as control variables. The forces and torques can then be synthesized, either in software or with dedicated hard-wired interfaces. Among the difficulties that occur when such maneuver techniques are employed is the fact that actuator saturation constraints are related to acceleration bounds in a state-dependent way. Recent work in the literature relies on generating a correction to the acceleration inputs, by pointwise constrained acceleration error minimization. This technique works best for infinite time horizons and highly coupled manipulator geometries, but not for terminal control or when the forces and torques enter the dynamics multiplied by a diagonal control influence matrix. An alternative technique discussed in the present paper consists of running actuators at saturation levels between sampling instants, then re-initializing an exact optimal regulator algorithm whenever the inputs drop below saturation range. Examples of such a maneuver strategy are given, both for asymptotic regulation and terminal control.