Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks266 from public data
- 6,702
- 1,902
- Collaborative Deep Learning for Recommender Systems1,727
A hierarchical Bayesian model called collaborative deep learning (CDL), which jointly performs deep representation learning for the content information and collaborative filtering for the ratings (feedback) matrix is proposed, which can significantly advance the state of the art.
- Learning a Deep Compact Image Representation for Visual Tracking974
Comparison with the state-of-the-art trackers on some challenging benchmark video sequences shows that the deep learning tracker is more accurate while maintaining low computational cost with real-time performance when the MATLAB implementation of the tracker is used with a modest graphics processing unit (GPU).
- Dynamic Key-Value Memory Networks for Knowledge Tracing933
This work introduces a new model called Dynamic Key-Value Memory Networks (DKVMN) that can exploit the relationships between underlying concepts and directly output a student's mastery level of each concept.
- The Visual Object Tracking VOT2016 Challenge Results776
The Visual Object Tracking challenge VOT2016 goes beyond its predecessors by introducing a new semi-automatic ground truth bounding box annotation methodology and extending the evaluation system with the no-reset experiment.
- 596
- Human Action Recognition Using Factorized Spatio-Temporal Convolutional Networks577
Factorized spatio-temporal convolutional networks (FstCN) are proposed that factorize the original 3D convolution kernel learning as a sequential process of learning 2D spatial kernels in the lower layers, followed by learning 1D temporal kernel in the upper layers.
- Robust path-based spectral clustering565
This paper develops a robust path-based spectral clustering method by defining a robust path-based similarity measure for spectral clustering under both unsupervised and semi-supervised settings and shows results that consistently outperforms other methods due to its higher robustness.
- Constructive algorithms for structure learning in feedforward neural networks for regression problems538
This survey paper first describes the general issues in constructive algorithms, with special emphasis on the search strategy, then presents a taxonomy, based on the differences in the state transition mapping, the training algorithm, and the network architecture.
- SVC2004: First International Signature Verification Competition503
The First International Signature Verification Competition (SVC2004) recently was organized as a step towards establishing common benchmark databases and benchmarking rules and the experience gained will be very useful to similar activities in the future.
- A convex formulation for learning task relationships in multi-task learning476
This paper proposes a regularization formulation for learning the relationships between tasks in multi-task learning, called MTRL, which can also describe negative task correlation and identify outlier tasks based on the same underlying principle.
- Visual Object Tracking for Unmanned Aerial Vehicles: A Benchmark and New Motion Models431
A simple baselines to model the camera motion by geometric transformation based on background feature points is devised to address the challenging issue of severe camera motion in an unmanned aerial vehicle (UAV) or drone scenario.
- Earthformer: Exploring Space-Time Transformers for Earth System Forecasting416
The idea is to decompose the data into cuboids and apply cuboid-level self-attention in parallel, which achieves state-of-the-art performance on real-world benchmarks about precipitation nowcasting and El Nino/Southern Oscillation forecasting.
- Host-based intrusion detection using dynamic and static behavioral models385
This paper adopts an anomaly detection approach by detecting possible intrusions based on program or user profiles built from normal usage data using a scheme that can be justified from the perspective of hypothesis testing.
- Multilingual and Multi-Aspect Hate Speech Analysis365
This paper presents a new multilingual multi-aspect hate speech analysis dataset and uses it to test the current state-of-the-art multilingual multitask learning approaches.
- Understanding and Diagnosing Visual Tracking Systems365
A framework by breaking a tracker down into five constituent parts, namely, motion model, feature extractor, observation model, model updater, and ensemble post-processor is proposed, which can provide a solid baseline when conducting controlled experiments for visual tracking research.
- Addressing two problems in deep knowledge tracing via prediction-consistent regularization363
This work introduces regularization terms that correspond to reconstruction and waviness to the loss function of the original DKT model to enhance the consistency in prediction and alleviates two major problems in the model.
- 337
- Mathematical expression recognition: a survey328
This survey paper will review most of the existing work with respect to each of the two major stages of the recognition process, and tries to put emphasis on the similarities and differences between systems.
- A Survey on Bayesian Deep Learning327
This survey provides a comprehensive introduction to Bayesian deep learning and reviews its recent applications on recommender systems, topic models, control, and so on.
- MagicDrive: Street View Generation with Diverse 3D Geometry Control310
This paper introduces MagicDrive, a novel street view generation framework, offering diverse 3D geometry controls including camera poses, road maps, and 3D bounding boxes, together with textual descriptions, achieved through tailored encoding strategies.
- 303
- 276
- 268
- Towards Bayesian Deep Learning: A Framework and Some Existing Methods262
This paper proposes a general framework for Bayesian deep learning and reviews its recent applications on recommender systems, topic models, and control and the relationship and differences between Bayesian deep learning and other related topics such as the Bayesian treatment of neural networks.
- RoboDreamer: Learning Compositional World Models for Robot Imagination239
RoboDreamer is introduced, an innovative approach for learning a compositional world model by factorizing the video generation, which can successfully synthesize video plans on unseen goals in the RT-X, enables successful robot execution in simulation, and substantially outperforms monolithic baseline approaches to video generation.
- 238
- 235
- 211
- Co-Regularized Hashing for Multimodal Data204
This paper proposes a novel multimodal hash function learning method, called Co-Regularized Hashing (CRH), based on a boosted co-regularization framework, and empirically compares CRH with two state-of-the-art multi-modal hash functions learning methods on two publicly available data sets.
- Adaptive Transfer Learning202
An Adaptive Transfer learning algorithm based on Gaussian Processes (AT-GP), which can be used to adapt the transfer learning schemes by automatically estimating the similarity between a source and a target task, is proposed.
- A probabilistic model for multimodal hash function learning201
This paper proposes a probabilistic model, called MLBE, to learn hash functions from multimodal data automatically, and devise an efficient algorithm for the learning of binary latent factors which corresponds to hash function learning.
- 196
- Spatiotemporal Modeling for Crowd Counting in Videos194
This work proposes a variant of a recent deep learning model called convolutional LSTM (ConvLSTM) for crowd counting which fully captures both spatial and temporal dependencies and extends the ConvLSTm model to a bidirectional ConvL STM model which can access long-range information in both directions.
- Robust locally linear embedding182
This paper proposes an approach to make LLE more robust, known as robust locally linear embedding (RLLE), and presents several specific methods for realizing this general RLLE approach.
- Overlapping community detection via bounded nonnegative matrix tri-factorization180
This paper proposes a method called bounded nonnegative matrix tri-factorization (BNMTF), which can explicitly model and learn the community membership of each node as well as the interaction among communities using three factors in the factorization.
- A Probabilistic Approach to Robust Matrix Factorization178
Experiments show that PRMF is comparable to other state-of-the-art robust matrix factorization methods in terms of accuracy and outperforms them particularly for large data matrices.
- TagiCoFi170
This paper proposes a novel framework, called tag informed collaborative filtering (TagiCoFi), to seamlessly integrate tagging information into the CF procedure, and demonstrates that TagiCoFi outperforms its counterpart which discards the tagging information even when it is available, and achieves state-of-the-art performance.
- Time series clustering with ARMA mixtures165
This paper proposes a model-based approach to the clustering of data patterns that are represented as sequences or time series possibly of difierent lengths using mixtures of autoregressive moving average (ARMA) models and derives an expectation-maximization (EM) algorithm for learning the mixing coe‐cients as well as the parameters of the component models.
- 164
- The United States COVID-19 Forecast Hub dataset159
The goal of this dataset is to establish a standardized and comparable set of short-term forecasts from modeling teams that can be used to develop ensemble models, communicate forecasts to the public, create visualizations, compare models, and inform policies regarding COVID-19 mitigation.
- Probing Toxic Content in Large Pre-Trained Language Models148
A method based on logistic regression classifiers is proposed to probe English, French, and Arabic PTLMs and quantify the potentially harmful content that they convey with respect to a set of templates to assess and mitigate the toxicity transmitted by PTL Ms.
- 147
- A Regularization Approach to Learning Task Relationships in Multitask Learning143
A regularization approach to learning the relationships between tasks in multitask learning that can also describe negative task correlation and identify outlier tasks based on the same underlying principle is proposed.
- 136
- 130
- Mixture of Cluster-Conditional LoRA Experts for Vision-Language Instruction Tuning128
A novel Mixture of Experts (MoE) architecture designed to activate task-customized model parameters based on instruction clusters is proposed, and a separate universal expert is further incorporated to improve generalization abilities of MoCLE for novel instructions.
- 123
- CLIP2: Contrastive Language-Image-Point Pretraining from Real-World Point Cloud Data116
This work proposes Contrastive Language-Image-Point Cloud Pretraining (CLIP2) to directly learn the transferable 3D point cloud representation in realistic scenarios with a novel proxy alignment mechanism, and exploits naturally-existed correspondences in 2D and 3D scenarios.
- Relation regularized matrix factorization112
A novel MF method called relation regularized matrix factorization (RRMF) is proposed, which seamlessly integrates both the relation information and the content information into a principled framework and a linear-time learning algorithm with convergence guarantee to learn the parameters of RRMF is proposed.
- 106
- Transfer metric learning by learning task relationships103
This paper proposes a convex formulation for multi-task metric learning by modeling the task relationships in the form of a task covariance matrix and adapts the formulation of multi- task metric learning to the transfer learning setting for their method, called TML.
- Tensor embedding methods99
This paper proposes some novel tensor embedding methods which, unlike previous methods, take data directly in the form of tensors of arbitrary order as input and allow the relationships between dimensions of a tensor representation to be efficiently characterized.
- Ensemble-Based Tracking: Aggregating Crowdsourced Structured Time Series Data98
This work proposes a factorial hidden Markov model (FHMM) for ensemble-based tracking by learning jointly the unknown trajectory of the target and the reliability of each tracker in the ensemble and devise a conditional particle filter algorithm by exploiting the structure of the joint posterior distribution of the hidden variables.
- Temporal Dynamic Graph LSTM for Action-Driven Video Object Detection96
This paper proposes a novel temporal dynamic graph Long Short-Term Memory network (TDGraph LSTM), which enables global temporal reasoning by constructing a dynamic graph that is based on temporal correlations of object proposals and spans the entire video.
- 96
- Probabilistic Multi-Task Feature Selection96
A probabilistic interpretation of the general multi-task feature selection problem using the l1,q norm is provided and a Probabilistic model is developed using the noninformative Jeffreys prior to extend the model to learn and exploit more general types of pairwise relationships between tasks.
- Error detection, error correction and performance evaluation in on-line mathematical expression recognition96
This paper incorporates an error detection and correction mechanism into a parser developed previously by us based on de"nite clause grammar (DCG), and proposes a performance evaluation scheme which can be used to demonstrate the e!ectiveness of both the symbol recognition and structural analysis stages by a single measure.
- Knowledge Query Network for Knowledge Tracing95
A novel concept called probabilistic skill similarity is introduced that relates the pairwise cosine and Euclidean distances between skill vectors to the odds ratios of the corresponding skills, which makes KQN interpretable and intuitive.
- 90
- 89
- Gaussian Shell Maps for Efficient 3D Human Generation83
This work introduces Gaussian Shell Maps (GSMs) as a framework that connects SOTA generator network archi-tectures with emerging 3D Gaussian rendering primitives using an articulable multi shell-based scaffold and demonstrates that GSMs suc-cessfully generate 3D humans when trained on single-view datasets, including SHHQ and DeepFashion.
- On-line handwritten alphanumeric character recognition using dominant points in strokes80
In this paper, an approach to on-line handwritten alphanumeric character recognition based on sequential handwriting signals is presented and the issue of reference (or template) set evolution is also addressed.
- Locally linear metric adaptation for semi-supervised clustering79
This paper proposes a new metric learning method which performs nonlinear transformation globally but linear transformation locally, and forms the learning problem as an optimization problem and presents two methods for solving it.
- 78
- 78
- Temporal Link Prediction: A Unified Framework, Taxonomy, and Review76
This survey provides a comprehensive review of existing TLP methods and proposes a unified encoder-decoder framework to formulate all the methods reviewed, where different approaches only differ in terms of some components of the framework.
- Semi-Supervised Multi-Task Regression76
This paper proposes a new supervised multi-task regression method called SMTR, which is based on Gaussian processes (GP) with the assumption that the kernel parameters for all tasks share a common prior, and incorporates unlabeled data into SMTR by changing the kernel function of the GP prior to a data-dependent kernel function, resulting in a semi-supervised extension, called SSMTR.
- 75
- 75
- Spectral Multimodal Hashing and Its Application to Multimedia Retrieval75
This paper proposes a new hashing-based method for fast multimodal multimedia retrieval based on spectral analysis of the correlation matrix of different modalities and develops an efficient algorithm that learns some parameters from the data distribution for obtaining the binary codes.
- 75
- Hidden-Mode Markov Decision Processes for Nonstationary Sequential Decision Making75
In sequential decision problems, Markov decision process (MDP) is a model formulation that has been commonly used, due to its generality, flexibility, and applicability to a wide range of problems.
- Multi-task Learning Using Generalized t Process71
A Bayesian extension is proposed by modeling the task covariance matrix as a random matrix with an inverse-Wishart prior and integrating it out to achieve Bayesian model averaging in a multi-task generalized t process (MTGTP).
- Detection Recovery in Online Multi-Object Tracking with Sparse Graph Tracker69
SGT converts video data into a graph where detections, their connections, and the relational features of two connected nodes are represented by nodes, edges, and edge features, respectively, which allow SGT to track targets with tracking candidates selected by top-K scored detections with large K.
- A Bayesian framework for deformable pattern recognition with application to handwritten character recognition67
This paper studies how to integrate deformable models into a Bayesian framework as a unified approach for modeling, matching, and classifying shapes and can achieve an accuracy of 94.7 percent with no rejection on a subset of handwritten digits from the NIST SD-1 dataset.
- Multi-Task Learning in Heterogeneous Feature Spaces65
A multi-task extension of linear discriminant analysis (LDA), called MTDA, is proposed, which can deal with learning tasks with different data representations and can alleviate the labeled data deficiency problem of LDA.
- Recognizing on-line handwritten alphanumeric characters through flexible structural matching65
This paper proposes a simple yet robust structural approach for recognizing on-line handwriting that is designed to achieve reasonable speed, fairly high accuracy and sufficient tolerance to variations, and maintains a high degree of reusability and hence facilitates extensibility.
- EMOVA: Empowering Language Models to See, Hear and Speak with Vivid Emotions64
This work proposes the EMOVA (EMotionally Omni-present Voice Assistant), to enable Large Language Models with end-to-end speech abilities while maintaining the leading vision-language performance and surprisingly notices that omni-modal alignment can further enhance vision-language and speech abilities compared with the bi-modal aligned counterparts.
- MultiSiam: Self-supervised Multi-instance Siamese Representation Learning for Autonomous Driving64
The Multi-instance Siamese Network (MultiSiam) remarkably improves generalization ability and achieves state-of-the-art transfer performance on autonomous driving benchmarks, while existing SSL counterparts like MoCo, MoCo-v2, and BYOL show significant performance drop.
- 63
- Relational Deep Learning: A Deep Latent Variable Model for Link Prediction63
This work follows the Bayesian deep learning framework and devise a hierarchical Bayesian model, called relational deep learning (RDL), to jointly model high-dimensional node attributes and link structures with layers of latent variables and derive a generalized variational inference algorithm for learning the variables and predicting the links.
- Mixed Autoencoder for Self-Supervised Visual Representation Learning62
The proposed Mixed Autoencoder (MixedAE) achieves the state-of-the-art transfer results among masked image modeling (MIM) augmentations on different downstream tasks with significant efficiency and is the very first work to consider mixing for MIM from the perspective of pretext task design.
- 61
- Multilabel relationship learning59
Through experiments conducted on some multilabel applications, it is found that MLRL not only gives higher classification accuracy but also has better interpretability as revealed by the label covariance matrix.
- Kernel-based distance metric learning for content-based image retrieval59
This paper proposes a kernel approach to improve the retrieval performance of CBIR systems by learning a distance metric based on pairwise constraints between images as supervisory information and defines the transformation in the kernel-induced feature space which is nonlinearly related to the image space.
- Parametric distance metric learning with label information57
A parametric method for metric learning based on class label information is proposed, which performs parametric learning to find a regression mapping from the input space to a feature space, such that the dissimilarity between patterns in the input space is approximated by the Euclidean distance between points in the feature space.
- An Environment Model for Nonstationary Reinforcement Learning53
This paper partially addresses the problem of reinforcement learning in nonstationary environments by formalizing a subclass of nonstationARY environments called hidden-mode Markov decision process (HM-MDP), which is a special case of partially observableMarkov decision processes (POMDP).
- Anyattack: Towards Large-scale Self-supervised Adversarial Attacks on Vision-language Models52
AnyAttack, a self-supervised framework that transcends the limitations of conventional attacks through a novel foundation model approach, is presented, enabling any image to be transformed into an attack vector targeting any desired output across different VLMs.
- 51
- Extending the relevant component analysis algorithm for metric learning using both positive and negative equivalence constraints51
This paper proposes an extension to RCA that allows both positive and negative equivalence constraints to be incorporated and shows that the extended RCA algorithm is effective.
- 51
- 51
- DetDiffusion: Synergizing Generative and Perceptive Models for Enhanced Data Generation and Perception49
DetDiffusion, for the first time, harmonizes generative and perceptive models, tackling the challenges in generating effective data for perceptive models, and introduces perception-aware loss through segmentation, improving both quality and controllability.
- Towards Bayesian Deep Learning: A Survey49
A general introduction to Bayesian deep learning is provided and its recent applications on recommender systems, topic models, and control are reviewed.
- A kernel path algorithm for support vector machines49
This paper learns the hyperparameter of the kernel function for a support vector machine (SVM) without having to train the model multiple times, and finds that the solutions of the neighborhood hyperparameters can be calculated exactly.
- Adaptive Online Replanning with Diffusion Models48
A principled approach to determine when to replan with diffusion models, based on the diffusion model's estimated likelihood of existing generated plans, to ensure that new plans follow the same goal state as the original trajectory, which may efficiently bootstrap off previously generated plans.
- 47
- A manifold regularization approach to calibration reduction for sensor-network based tracking47
A manifold regularization approach known as LeMan to calibration-effort reduction for tracking a mobile node in a wireless sensor network and compute a subspace mapping function between the signal space and the physical space by using a small amount of labeled data and a large amount of unlabeled data.
- Locally linear metric adaptation with application to semi-supervised clustering and image retrieval46
Experimental results based on two real-world image databases show that LLMA significantly outperforms other methods in boosting the image retrieval performance and is used to improve the performance of content-based image retrieval systems through metric learning.
- Worst-Case Linear Discriminant Analysis45
A new dimensionality reduction method called worst-case linear discriminant analysis (WLDA) is proposed by defining new between-class and within-class scatter measures and adopts the worst- case view which arguably is more suitable for applications such as classification.
- 45
- 44
- 42
- Multi-Task Boosting by Exploiting Task Relationships42
This paper proposes a method called MTBoost, which generalizes boosting to the multi-task learning setting and conducts theoretical analysis on the convergence of MTBoost and also empirical analysis comparing it with several related methods.
- 42
- Learning Unmanned Aerial Vehicle Control for Autonomous Target Following41
A hierarchical approach is developed that combines a model-free policy gradient method with a conventional feedback proportional-integral-derivative (PID) controller to enable stable learning without catastrophic failure and can learn a target following policy in a simulator efficiently.
- Incorporating Features Learned by an Enhanced Deep Knowledge Tracing Model for STEM/Non-STEM Job Prediction40
Detailed analysis on the student’s knowledge state reveals that, when compared with non-STEM students, STEM students generally show a higher mastery level and a higher learning gain in mathematics.
- Surrogate maximization/minimization algorithms and extensions40
The usefulness of SM algorithms is demonstrated by taking logistic regression models, AdaBoost and the log-linear model as examples and devise several SM algorithms, including the standard SM, generalized SM, gradient SM, and quadratic SM algorithms.
- 39
- 3D-Aware Indoor Scene Synthesis with Depth Priors39
This work proposes a dual-path generator, where one path is responsible for depth generation, whose intermediate features are injected into the other path as the condition for appearance rendering, which eases the 3D-aware synthesis with explicit geometry information.
- Two-dimensional solution path for support vector regression38
This paper shows that the solution path for ε-SVR is also piecewise linear with respect to ε, and proposes an efficient algorithm for exploring the two-dimensional solution space defined by the regularization and error parameters.
- Structuring Visual Words in 3D for Arbitrary-View Object Localization37
A novel and efficient method for generic arbitrary-view object class detection and localization that can automatically determine the locations and outlines of multiple objects in the test image with occlusion handling and can accurately estimate both the intrinsic and extrinsic camera parameters in an optimized way.
- Use of bias term in projection pursuit learning improves approximation and convergence properties37
It is shown that by including a bias term in each linear projection of the predictor variables, PPL networks can regain these capabilities, independent of the exact choice of R, and is applied to chaotic time series prediction.
- 35
- 35
- Probabilistic Graphical Models for Boosting Cardinal and Ordinal Peer Grading in MOOCs34
This paper proposes novel extensions to some existing probabilistic graphical models for peer grading that give superior performance in cardinal evaluation and also outperform conventional ordinal models in ordinal eval- uation.
- Face recognition using a kernel fractional-step discriminant analysis algorithm34
This paper proposes an enhanced kernel discriminant analysis (KDA) algorithm called kernel fractional-step discriminantAnalysis (KFDA) for nonlinear feature extraction and dimensionality reduction and shows that this new algorithm outperforms traditional kernel principal component analysis ( KPCA) and KDA algorithms.
- 33
- Transfer Metric Learning with Semi-Supervised Extension33
This article considers a transfer-learning setting in which some related source tasks with labeled data are available to help the learning of the target task, and proposes a semi-supervised extension of TML called STML to further improve the generalization performance by exploiting the unlabeled data based on the manifold assumption.
- 32
- Gaussian Process Latent Random Field32
A novel supervised extension of GPLVM is proposed, called Gaussian process latent random field (GPLRF), by enforcing the latent variables to be a Gaussian Markov random field with respect to a graph constructed from the supervisory information.
- Bayesian regularization in constructive neural networks30
Simulation shows that regularization, with input training using a full Bayesian approach, produces networks with better generalization performance and lower susceptibility to over-fitting as the network size increases.
- Surrogate maximization/minimization algorithms for AdaBoost and the logistic regression model29
This paper solves the boosting problem by proposing SM algorithms for the corresponding optimization problem and derives an SM algorithm that can be shown to be identical to the algorithm proposed by Collins et al. (2002) based on Bregman distance.
- Gesture-based piloting of an aerial robot using monocular vision28
This work focuses on building an economical monocular system particularly designed for gesture-based piloting of an aerial robot, and natural arm gestures are mapped to rich target directions and convenient fine adjustment is achieved.
- Bidirectional deformable matching with application to handwritten character extraction28
This paper describes two Bayesian frameworks, one for classifying input patterns and another for detecting target patterns in complex scenes using deformable models, and shows that the two frameworks are similar to the forward-reverse setting of Hausdorff matching and that their matching and discriminating properties are complementary to each other.
- Comparative Evaluation of Label-Agnostic Selection Bias in Multilingual Hate Speech Datasets27
This work uses topic models to discover latent semantics in eleven hate speech corpora, then presents two bias evaluation metrics based on the semantic similarity between topics and search words frequently used to build corpora.
- Learning the kernel matrix by maximizing a KFD-based class separability criterion27
This paper proposes a novel method for learning the kernel matrix based on maximizing a class separability criterion that is similar to those used by linear discriminant analysis (LDA) and kernel Fisher discriminant (KFD).
- Implicit Concept Removal of Diffusion Models26
The Geom-Erasing effectively mitigates the generation of implicit concepts, achieving the state-of-the-art results on the Inappropriate Image Prompts (I2P) and the authors' challenging Implicit Concept Dataset (ICD) benchmarks.
- A Heuristic Search Approach to Chinese Glyph Generation Using Hierarchical Character Composition26
This work proposes structurally based character composition expressions (CCE) for com - posing these unrepresented Chinese characters unambiguously, and describes a glyph gen- eration process, which generates the corresponding glyph for a given CCE.
- 25
- Learning 3D-Aware Image Synthesis with Unknown Pose Distribution25
This work proposes PoF3D, a pose-free generator and the pose-aware discriminator that frees generative radiance fields from the requirements of 3D pose priors, and demonstrates the feasibility of learning high-quality 3D-aware image synthesis without using 3D posing priors for the first time.
- 25
- 25
- Relaxational metric adaptation and its application to semi-supervised clustering and content-based image retrieval25
A nonparametric method, called relaxational metric adaptation (RMA), for the same metric adaptation problem, which is local in the sense that it allows locally adaptive metrics, and global because even patterns not in the vicinity can have long-range effects on the metric adaptation process.
- 25
- Towards a Better Tradeoff between Quality and Efficiency of Community Detection: An Inductive Embedding Method across Graphs24
This study considers the CD task and proposes a novel inductive community detection method, which can achieve a significant tradeoff between quality and efficiency over various baselines, and develops a novel feature extraction module based on graph coarsening which can efficiently extract informative feature inputs for GNNs.
- Point-Cloud-Based Place Recognition Using CNN Feature Extraction24
A novel point-cloud-based place recognition system that adopts a deep learning approach for feature extraction by using a convolutional neural network pre-trained on color images to extract features from a range image without fine-tuning on extra range images is proposed.
- Active hashing and its application to image and text retrieval24
This paper proposes a novel framework, called active hashing, which can actively select the most informative labeled pairs for hash function learning and develops a batch mode algorithm to speed up active selection.
- Constructive neural networks as estimators of bayesian discriminant functions24
It can be shown that the resultant network approximates a Bayesian classifier that implements the Bayesian decision rule for classification.
- 23
- Probabilistic Relational PCA23
By explicitly modeling covariance between instances as derived from the relational information, a novel probabilistic dimensionality reduction method is proposed, called Probabilistic relational PCA (PRPCA), for relational data analysis.
- Model-based transductive learning of the kernel matrix23
This paper addresses the problem of transductive learning of the kernel matrix from a probabilistic perspective and devise an expectation-maximization (EM) algorithm to infer the missing data, parameter matrix and feature dimensionality in a maximum a posteriori (MAP) manner.
- Effective Feature Learning with Unsupervised Learning for Improving the Predictive Models in Massive Open Online Courses22
This paper makes the first attempt to solve the feature learning problem by taking the unsupervised learning approach to learn a compact representation of the raw features with a large degree of redundancy.
- 22
- Semi-supervised Discriminant Analysis Via CCCP22
A novel semi-supervised discriminant analysis algorithm called SSDA, utilize unlabeled data to maximize an optimality criterion of LDA and use the constrained concave-convex procedure to solve the optimization problem.
- 22
- 22
- 21
- 21
- A Convex Formulation of Modularity Maximization for Community Detection21
This paper proposes a reformulation of an optimization problem for the 2-partition problem with a convex relaxation scheme to give an iterative algorithm which solves a simple quadratic program in each iteration and finds that this method is both scalable and competitive in performance via maintaining a good tradeoff between efficiency and quality.
- A Neural Network Approach to Constructive Induction21
It is argued that the learning process takes advantage of the transfer effects of prior learning in constructing large networks from smaller ones, which is a dynamic network construction process which involves adjusting the network weights as well as the topology.
- Bayesian inference for transductive learning of kernel matrix using the Tanner-Wong data augmentation algorithm20
By regarding the transductive learning of the kernel matrix as a missing data problem, a Bayesian hierarchical model is proposed and the Tanner-Wong data augmentation algorithm for making inference on the model is devised.
- SVQNet: Sparse Voxel-Adjacent Query Network for 4D Spatio-Temporal LiDAR Semantic Segmentation19
The Sparse Voxel-Adjacent Query Network (SVQNet) for 4D LiDAR semantic segmentation achieves state-of-the-art performance in LiDAR semantic segmentation of the SemanticKITTI benchmark and the nuScenes dataset.
- 19
- 19
- 19
- On reducing learning time in context-dependent mappings19
An approach to overcoming the slow convergence problems often associated with learning complex nonlinear mappings in a context-dependent manner so that complex problems are decomposed into simpler subproblems corresponding to different contexts.
- Kernel selection forl semi-supervised kernel machines18
This paper proposes an integrated regularization framework for semi-supervised kernel machines by incorporating both the cluster assumption and the manifold assumption, which supports kernel learning in the form of kernel selection.
- Bayesian inference on principal component analysis using reversible jump Markov chain Monte Carlo18
This work presents a hierarchical model for probabilistic PCA and constructs a Bayesian inference method for this model using reversible jump Markov chain Monte Carlo (MCMC), capable of automatically determining the number of principal components and estimating the parameters simultaneously under the same disciplined framework.
- A Scalable Kernel-Based Semisupervised Metric Learning Algorithm with Out-of-Sample Generalization Ability17
This letter proposes a nonlinear metric learning method based on the kernel approach that can handle significantly larger data sets and can naturally lead to out-of-sample generalization.
- Throttling spoofed SYN flooding traffic at the source17
A simple and efficient method to detect and defend against TCP SYN flooding attacks under different IP spoofing types, including subnet spoofing, which makes use of a storage-efficient data structure and a change-point detection method to distinguish complete three-way TCP handshakes from incomplete ones.
- User Profiling for Intrusion Detection Using Dynamic and Static Behavioral Models17
This paper adopts an anomaly detection approach by detecting possible intrusions based on user profiles built from normal usage data using a scheme that can be justified from the perspective of hypothesis testing and results show that static modeling outperforms dynamic modeling for this application.
- 16
- Fine-grained categorization via CNN-based automatic extraction and integration of object-level and part-level features15
By extracting and interpreting the hierarchical hidden layer features learned by a CNN, an elaborate CNN-based system for fine-grained categorization is proposed which achieves impressive performance that is superior or comparable to the state of the art.
- Sparse Boltzmann Machines with Structure Learning as Applied to Text Analysis15
A method for learning Sparse Boltzmann Machines, where each hidden unit is connected to a subset of the visible units instead of all of them, which yields models with significantly improved model fit and interpretability as compared with RBMs.
- 14
- Heteroscedastic Probabilistic Linear Discriminant Analysis with Semi-supervised Extension14
A semi-supervised and heteroscedastic extension of probabilistic LDA, called S2HPLDA, which aims at overcoming all limitations under a common principled framework and applies automatic relevance determination to determine the required dimensionality of the low-dimensional space for dimensionality reduction.
- Learning Two-View Stereo Matching14
A graph-based semi-supervised symmetric matching framework that performs dense matching between two uncalibrated wide-baseline images by exploiting the results of sparse matching as labeled data that is robust against outliers in the initial sparse matching.
- Stepwise Metric Adaptation Based on Semi-Supervised Learning for Boosting Image Retrieval Performance14
This paper proposes a stepwise LLMA algorithm for boosting the retrieval performance of CBIR systems by incorporating relevance feedback from users collected over multiple query sessions, and reports experimental results performed on a real-world color image database.
- 13
- Posterior Consistency of the Silverman g-prior in Bayesian Model Choice13
A Bayesian interpretation of sparsity in the kernel setting is pursued by making use of a mixture of a point-mass distribution and prior that is referred to as "Silverman's g-prior", which provides a theoretical analysis of the posterior consistency of a Bayesian model choice procedure based on this prior.
- 13
- 13
- Clickstream Knowledge Tracing: Modeling How Students Answer Interactive Online Questions12
This paper proposes the first clickstream-based KT model, dubbed clickstream knowledge tracing (CKT), which augments a basic KT model by modeling the clickstream activities of students when answering questions and shows that incorporating clickstream data leads to performance improvement.
- Kernel-Based Metric Adaptation with Pairwise Constraints12
A kernel-based approach for nonlinear metric learning is proposed, which performs locally linear translation in the kernel-induced feature space by solving the metric learning problem efficiently by kernel matrix adaptation.
- 12
- SitEmb-v1.5: Improved Context-Aware Dense Retrieval for Semantic Association and Long Story Comprehension11
It is shown that existing embedding models are not well-equipped to encode such situated context effectively, and thus a new training paradigm is introduced and the situated embedding models (SitEmb) are developed.
- Unified Triplet-Level Hallucination Evaluation for Large Vision-Language Models11
A unified framework to measure the object and relation hallucination in LVLMs simultaneously and introduces Tri-HE, a novel Triplet-level Hallucination Evaluation benchmark which can be used to study both object and relation hallucination at the same time.
- Stereo Waterdrop Removal with Row-wise Dilated Attention11
A novel row-wise dilated attention module is proposed to enlarge attention’s receptive field for effective information propagation between the two stereo images to better detect and remove waterdrops from stereo images.
- 11
- 11
- Pre-train and Refine: Towards Higher Efficiency in K-Agnostic Community Detection without Quality Degradation10
This study proposes PRoCD (Pre-training & Refinement fOr Community Detection), a simple yet effective method that reformulates K-agnostic CD as the binary node pair classification and demonstrates that PRoCD can ensure higher efficiency in K-agnostic CD without significant quality degradation.
- SED10
This paper proposes a novel active learning method for text classification, called supervised experimental design (SED), which seamlessly incorporates label information into experimental design and shows that SED outperforms its counterparts which either discard the label information even when it is available or fail to exploit the distribution of unlabeled data.
- On-line handwritten alphanumeric character recognition using feature sequences10
This paper presents an approach in which an on-line handwritten character is characterized by a sequence of dominant points in strokes and a sequences of writing directions between consecutive dominant points.
- 9
- Coherence Functions for Multicategory Margin-based Classification Methods9
A new majorization loss function is proposed that is closely related to the multinomial log-likelihood function and its limit at zero temperature corresponds to a multicategory hinge loss function.
- 9
- 9
- 8
- ECCV 2024 W-CODA: 1st Workshop on Multimodal Perception and Comprehension of Corner Cases in Autonomous Driving8
Details of the 1st W-CODA workshop are presented, held in conjunction with the ECCV 2024, to explore next-generation solutions for autonomous driving corner cases, empowered by state-of-the-art multimodal perception and comprehension techniques.
- Transductive Learning on Adaptive Graphs8
By incorporating label information, this work seeks to enhance the performance of graph-based semi-supervised learning by learning the graph and label inference simultaneously by proposing an iterative algorithm to solve the optimization problem which has closed-form solution in each step.
- Boosting kernel discriminant analysis and its application to tissue classification of gene expression data8
A new algorithm to further improve the overall performance of KDA by effectively integrating the boosting and KDA techniques is developed, which constitutes a strong ensemble-based KDA framework.
- Extending Kernel Fisher Discriminant Analysis with the Weighted Pairwise Chernoff Criterion8
A new KFD method called heteroscedastic kernel weighted discriminant analysis (HKWDA) is proposed which has several appealing characteristics and can handle nonlinearity efficiently in a disciplined manner and effectively deal with the small sample size problem.
- 7
- Controlled Text Generation Using Dictionary Prior in Variational Autoencoders7
Dictionary Prior is proposed, a new data-driven prior that enjoys the merits of expressivity and controllability and is proposed to employ contrastive learning to separate the latent space into several parts to facilitate controlled text generation with DPrior.
- Social Relations Model for Collaborative Filtering7
A novel probabilistic model for collaborative filtering (CF), called SRMCoFi, which seamlessly integrates both linear and bilinear random effects into a principled framework, which also integrates their advantages while simultaneously overcoming their disadvantages.
- A scalable kernel-based algorithm for semi-supervised metric learning7
This paper proposes a nonlinear metric learning method based on the kernel approach that can handle significantly larger data sets and can naturally lead to out-of-sample generalization.
- 7
- Solving hidden-mode Markov decision problems7
This paper exploits the HM-MDP structure and derives an equation for dynamic programming update and substantially reduces the amount of computations in (cid:12)nding optimal policies, and shows that the HM-MDP approach outperforms the POMDP one several order of magnitude with respect to both space requirement and speed.
- 6
- 6
- High-level LoRA and hierarchical fusion for enhanced micro-expression recognition6
HLoRA-MER is a novel framework that combines high-level low-rank adaptation (HLoRA) and a hierarchical fusion module (HFM) that fine-tunes the high-level layers of a VFM to capture facial muscle movement information, while HFM aggregates inter-frame and spatio-temporal features.
- 6
- Instance-specific canonical correlation analysis6
A novel algorithm called Instance-Specific Canonical Correlation Analysis (ISCCA), which approximates the nonlinear data by computing the instance-specific projections along the smooth curve of the manifold and model locality and nonlinearity jointly for multi-view correlation learning.
- Multicategory large margin classification methods: Hinge losses vs. coherence functions6
The Fisher consistency properties of multicategorymajorization losses are explored and a construction framework of majorization losses of the 0-1 loss is presented and proposed based on a coherence function.
- Semisupervised metric learning by kernel matrix adaptation6
A kernel-based approach for nonlinear metric learning, which performs locally linear translation in the kernel-induced feature space by kernel matrix adaptation, which is promising for semi-supervised metric learning.
- 6
- 6
- 6
- Micro-Expression Recognition via Fine-Grained Dynamic Perception5
This article proposes to rank frame-level features of a sequence of raw frames in chronological order, in which the rank process encodes the dynamic information of both ME appearances and motions, and develops a novel fine-grained dynamic perception (FDP) framework for MER.
- Learning Accurate Objectness Instance Segmentation from Photorealistic Rendering for Robotic Manipulation5
This paper develops an automated rendering pipeline for rapidly generating labeled datasets and develops an RGB-D segmentation model by extending the Mask R-CNN framework for depth input fusion.
- Towards efficient structural analysis of mathematical expressions5
This paper uses Definite Clause Grammar (DCG) as a formalism to define a set of replacement rules for parsing mathematical expressions, and proposes some methods here to increase the efficiency of the parsing process.
- STLDM: Spatio-Temporal Latent Diffusion Model for Precipitation Nowcasting4
STLDM is a diffusion-based model that learns the latent representation from end to end alongside both the Variational Autoencoder and the conditioning network, and achieves superior performance compared to the state of the art, while also improving inference efficiency.
- 4
- YANG Data Model for Key Chains4
This document describes the key chain YANG data model, a list containing one or more elements containing a Key ID, key string, send/accept lifetimes, and the associated authentication or encryption algorithm.
- A regularization framework for multiclass classification: A deterministic annealing approach4
A general regularization framework for multiclass classification based on discriminant functions is proposed, and class-conditional posterior probabilities can be calculated directly without assuming the underlying probabilistic model.
- Sliced Coordinate Analysis for Effective Dimension Reduction and Nonlinear Extensions4
A notion of nonlinear effective dimension reduction is proposed and a nonlinear extension of SIR is developed called kernel SIR (KSIR), which are based on principal component analysis and sliced coordinate analysis, respectively.
- 4
- 4
- Many-Shot CoT-ICL: Making In-Context Learning Truly Learn3
This work studies many-shot ICL on reasoning tasks, with a particular focus on many-shot chain-of-thought in-context learning (CoT-ICL), and identifies several distinctive properties of many-shot CoT-ICL.
- 3
- 3
- 3
- Rethinking Targeted Adversarial Attacks for Neural Machine Translation3
Experimental results demonstrate that the proposed setting for NMT targeted adversarial attacks could provide faithful attacking results for targeted adversarial attacks on NMT systems, and the proposed TWGA method can effectively attack such victim NMT systems.
- 3
- Semiparametric preference learning3
A semiparametric preference learning model, abbreviated as SPPL, with the aim of combining the strengths of the parametric and nonparametric approaches is proposed, which is more powerful than previous models while keeping the computational complexity low.
- Discriminative Experimental Design3
This paper extends a previous active learning method called transductive experimental design by proposing a new unlabeled data selection criterion, called discriminative experimental design (DED), which incorporates both margin-based discrim inative information and data distribution information and hence it can be seen as a discrimInative extension of TED.
- Semi-supervised Cast Indexing for Feature-Length Films3
A novel cast indexing approach based on hierarchical clustering, semi-supervised learning and linear discriminant analysis of the facial images appearing in the video sequence that can efficiently discover the main characters in such videos and retrieve their associated shots.
- 3
- Multi-model approach to non-stationary reinforcement learning3
This paper proposes a novel alogrithm for a class of nonstationary reinforcement learning problems in which the environmental changes are rare and finite, and consistently outperforms the finite window approach in various empirical setups.
- A SIMPLE YET ROBUST STRUCTURAL APPROACH FOR ON-LINE HANDWRITTEN ALPHANUMERIC CHARACTER RECOGNITION3
A simple yet robust structural approach for recognizing on-line handwriting designed to achieve reasonable speed, fairly high accuracy and suucient tolerance to variations, which maintains a high degree of reusability and hence facilitates extensibility.
- 2
- An online forecasting-based fine-tuning pipeline for time-series anomaly prediction2
A novel exemplar-based pre-training and fine-tuning pipeline tailored to time-series anomaly detection, which proves the proposed method's effectiveness in addressing the new TSAP tasks, which are incapable of being handled by current time-series anomaly detection or online time-series forecasting methods.
- 2
- TransformMix: Learning Transformation and Mixing Strategies from Data2
An automated approach to learn better transformation and mixing augmentation strategies from data, TransformMix, which applies learned transformations and mixing masks to create compelling mixed images that contain correct and important information for the target tasks.
- IRWE: Inductive Random Walk for Joint Inference of Identity and Position Network Embedding2
An inductive random walk embedding (IRWE) method is proposed, which combines multiple attention units to handle the random walk on graph topology and simultaneously derives identity and position embeddings that are jointly optimized.
- Trading off Quality for Efficiency of Community Detection: An Inductive Method across Graphs2
This work proposes an alternative inductive community detection (ICD) method across graphs of a system or scenario to alleviate the NP-hard challenge and demonstrates that ICD can achieve a significant trade-off between quality and efficiency over various baselines.
- Supervised Probabilistic Robust Embedding with Sparse Noise2
This paper proposes a supervised probabilistic robust embedding (SPRE) model in which data are corrupted either by sparse noise or by a combination of Gaussian and sparse noises and devise a two-fold variational EM learning algorithm in which the update of model parameters has analytical solution.
- 2
- 2
- 2
- 2
- Theoretical aspects of neural computation : a multidisciplinary perspective : International Workshop, TANC '97, Hong Kong, 26-28 May 19972
This work contains the revised and edited version of the papers presented at the Hong Kong International Workshop on Theoretical Neural Computation: A Multidisciplinary Perspective, held in May 1997 in Hong Kong.
- Improving the approximation and convergence capabilities of projection pursuit learning2
It is demonstrated that PPL networks do not have the universal approximation and strong convergence properties for any finiteR, but, by including a bias term in each linear combination of the predictor variables, P PL networks can regain these capabilities, independent of the exact choice ofR.
- 1
- 1
- SCAT: Robust Self-supervised Contrastive Learning via Adversarial Training for Text Classification1
A novel learning framework called SCAT (Self-supervised Contrastive Learning via Adversarial Training), which can learn robust representations without requiring labeled data and can also significantly improve the robustness of existing pre-trained language models.
- Sparse Probabilistic Relational Projection1
A novel model, called sparse probabilistic relational projection (SPRP), to learn a sparse projection matrix for relational dimensionality reduction by imposing on the projection matrix a sparsity-inducing prior such as the Laplace prior or Jeffreys prior is proposed.
- 1
- Keynote: Bayesian Deep Learning Models for Recommendation Applications1
In this talk, some BDL models which have been proposed recently for recommendation applications will be presented and some follow-up work and outstanding research issues for B DL models will also be discussed.
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- Deep COVID-19 Forecasting for Multiple States with Data Augmentation–
To overcome the issue of scarcity of training data for this COVID-19 pandemic, a novel data augmentation technique is developed to generate useful data for training and the generated data can also be used for model validation.
- Semi-semantic Line-Cluster Assisted Monocular SLAM for Indoor Environments–
This paper presents a novel method to reduce the scale drift for indoor monocular simultaneous localization and mapping (SLAM) using the prior knowledge that in the indoor environment, the line segments form tight clusters.
- –
- Some Research Problems in Metric Learning and Manifold Learning–
This study focuses on semisupervised learning in the metric learning context, which aims to automatically discover the low-dimensional nonlinear manifold in a high-dimensional data space and then embed the data points into a low- dimensional embedding space.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.