Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks222 from public data
- The Sixth Visual Object Tracking VOT2018 Challenge Results804
The Visual Object Tracking challenge VOT2018 is the sixth annual tracker benchmarking activity organized by the VOT initiative; results of over eighty trackers are presented; many are state-of-the-art trackers published at major computer vision conferences or in journals in the recent years.
- A Comprehensive Overhaul of Feature Distillation788
A novel feature distillation method in which the distillation loss is designed to make a synergy among various aspects: teacher transform, student transform, distillation feature position and distance function, which achieves a significant performance improvement in all tasks.
- 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.
- Knowledge Transfer via Distillation of Activation Boundaries Formed by Hidden Neurons637
This paper proposes a knowledge transfer method via distillation of activation boundaries formed by hidden neurons and proposes an activation transfer loss that has the minimum value when the boundaries generated by the student coincide with those by the teacher.
- The Visual Object Tracking VOT2014 Challenge Results588
The Visual Object Tracking challenge 2014, VOT2014, aims at comparing short-term single-object visual trackers that do not apply pre-learned models of object appearance and introduces a new dataset with full annotation of targets by rotated bounding boxes and per-frame attribute.
- 530
- 311
- Symmetric Graph Convolutional Autoencoder for Unsupervised Graph Representation Learning291
A new numerically stable form of the Laplacian sharpening by incorporating the signed graphs is proposed and a new cost function which finds a latent representation and a latent affinity matrix simultaneously is devised to boost the performance of image clustering tasks.
- The Majority Can Help the Minority: Context-rich Minority Oversampling for Long-tailed Classification238
A novel minority over-sampling method to augment di-versified minority samples by leveraging the rich context of the majority classes as background images, which achieves state-of-the-art performance on various long-tailed classification benchmarks.
- Context-Aware Deep Feature Compression for High-Speed Visual Tracking229
A new context-aware correlation filter based tracking framework to achieve both high computational speed and state-of-the-art performance among real-time trackers and introduces extrinsic denoising processes and a new orthogonality loss term for pre-training and fine-tuning of the expert autoencoders.
- Influence-Balanced Loss for Imbalanced Visual Classification198
A new loss is derived used in the balancing training phase that alleviates the influence of samples that cause an overfitted decision boundary and outperforms the state-of-the-art cost-sensitive loss methods.
- 189
- Intelligent visual surveillance — A survey181
The first part surveys image enhancement, moving object detection and tracking, and motion behavior understanding, and the second part reviews wide-area surveillance techniques based on the fusion of multiple visual sensors, camera calibration and cooperative camera systems.
- Sensitivity analysis of multilayer perceptron with differentiable activation functions180
A sensitivity depending on the weight set in a single-output multilayer perceptron (MLP) with differentiable activation functions is proposed and computer simulations have been performed, resulting in good agreement between theoretical and simulation outcomes for small weight perturbations.
- Knowledge Distillation with Adversarial Samples Supporting Decision Boundary163
A new perspective based on a decision boundary is provided, which is one of the most important component of a classifier, and the proposed algorithm trains a student classifier based on the adversarial samples supporting the decision boundary.
- Flux observer with online tuning of stator and rotor resistances for induction motors124
Simulations and actual experiments show that the rotor flux is estimated in all operating conditions and that both resistances converge to their actual values when the electrical motor torque exists.
- 118
- VisDrone-SOT2018: The Vision Meets Drone Single-Object Tracking Challenge Results87
The evaluation protocol of the VisDrone-SOT2018 challenge is presented and the results of a comparison of 22 trackers on the benchmark dataset are presented, which are publicly available on the challenge website.
- Detection of moving objects with a moving camera using non-panoramic background model86
A small-size background model is built by the proposed single spatio-temporal distributed Gaussian model, whose size is the same as input frame, to decrease computation time and memory storage without loss of detection performance.
- 80
- Nonlinear Adaptive Control Using Networks of76
A stable nonparametric adaptive control approach using a piecewise local linear approximator with universal approximation capability and analyses of asymptotic stability of the tracking error and parameter convergence for the proposed adaptive control scheme with the on-line self-organizing structure are presented.
- Nonlinear adaptive control using networks of piecewise linear approximators76
A stable nonparametric adaptive control approach using a piecewise local linear approximator with universal approximation capability and analyses of asymptotic stability of the tracking error and parameter convergence for the proposed adaptive control scheme with the on-line self-organizing structure are presented.
- PMnet: learning of disentangled pose and movement for unsupervised motion retargeting72
The proposed deep learning framework for unsupervised motion retargeting outperforms the state-of-the-art (sota) method by reducing the motion retargeting error from 7.68 (sota) to 1.95 (the authors').
- Computer-aided Prostate Cancer Detection using Texture Features and Clinical Features in Ultrasound Image65
A new prostate detection method using multiresolution autocorrelation texture features and clinical features such as location and shape of tumor, which can detect cancerous tissues efficiently and high sensitivity by the measurement of the number of correctly classified pixels.
- 63
- 50
- A learning approach of wafer temperature control in a rapid thermal processing system50
This paper presents a learning approach for wafer temperature control in a rapid thermal processing system (RTP) without exact information on the dynamics, and it is demonstrated that the proposed method can achieve an accurate output tracking even without an exact RTP model.
- 48
- Skeleton-Based Action Recognition of People Handling Objects47
A new framework for recognizing object-related human actions by graph convolutional networks using human and object poses is proposed, which outperforms the state-of-the-art method for skeleton-based action recognition.
- 47
- AutoLR: Layer-wise Pruning and Auto-tuning of Learning Rates in Fine-tuning of Deep Networks46
It is discussed that trends of layer-wise weight variations by fine-tuning using a single learning rate do not match the well-known notion that lower- level layers extract general features and higher-level layers extract specific features.
- 45
- Two-stage online inference model for traffic pattern analysis and anomaly detection45
A method for modeling trajectory patterns with both regional and velocity observations through the probabilistic topic model by embedding Gaussian models into the discrete topic model framework, which uses continuous velocity as well as regional observations unlike existing approaches.
- 44
- Scene conditional background update for moving object detection in a moving camera43
A moving object detection algorithm adapting to various scene changes in a moving camera that adapts itself to dynamic scene changes and outperforms state-of-the-art methods.
- Hierarchical Kalman-particle filter with adaptation to motion changes for object tracking43
A new tracking method with improved efficiency and accuracy based on the subspace representation and particle filter, and a more physically meaningful proposal distribution of the particle filter with consideration of the nature of motion is presented.
- Theoretical analysis on feature extraction capability of class-augmented PCA43
The theoretical analysis aims to clarify the role of processes for encoding the class information, augmenting the encoded information to data, and extracting features from class-augmented data by applying PCA and to provide an explanation on how CA-PCA can extract good features.
- 43
- Gaussian-response correlation filter for robust visual object tracking40
This paper develops a correlation filter for predicting the target location based on the distribution of correlation response, and proposes an adaptive scale estimation method to detect the target scale changes during the tracking.
- 39
- 37
- Variational Autoencoded Regression: High Dimensional Regression of Visual Data on Complex Manifold35
This paper proposes a new high dimensional regression method by merging Gaussian process regression into a variational autoencoder framework, and introduces a strategy to learn the latent space as well as the encoder and decoder so that the result of the regressed responded coincide with the corresponding response in the data space.
- Unsupervised Hyperbolic Representation Learning via Message Passing Auto-Encoders33
A novel hyperbolic message passing auto-encoder whose overallAuto-encoding is performed inhyperbolic space and is validated through extensive quantitative and qualitative analyses to validate the properties and benefits of the unsupervisedHyperbolic representations.
- Adaptive control for output feedback nonlinear systems in the presence of modeling errors33
The paper is extending output feedback nonlinear control and backstepping approaches to a class of systems approximately diffeomorphic to output feedback systems that include unknown functions.
- Context-Based Parking Slot Detection With a Realistic Dataset31
This paper proposes a context-based parking slot detection method inspired by the process of a human driver finding a parking slot that outperformed the previous deep-learning-based methods, along with having a short operation time.
- Gaussian Mixture Proposals with Pull-Push Learning Scheme to Capture Diverse Events for Weakly Supervised Temporal Video Grounding30
To learn moderately coupled Gaussian mixture capturing diverse events, this work newly proposes a pull-push learning scheme using pulling and pushing losses, each of which plays an opposite role to the other.
- 30
- 29
- 29
- 28
- Linear boundary discriminant analysis28
A new discriminantAnalysis, called linear boundary discriminant analysis (LBDA), which increases class separability by reflecting the different significances of non-boundary and boundary patterns by defining two novel scatter matrices and solving the eigenproblem on the criterion described by these scatterMatrices.
- 27
- A design framework for hierarchical ensemble of multiple feature extractors and multiple classifiers26
A hierarchical ensemble of MFMC and its optimizing framework is suggested that integrates reinforcement machine learning and Bayesian network and outperforms the previous ensemble methods in terms of accuracy.
- 26
- 25
- Domain density description for multiclass pattern classification with reduced computational load24
Experimental results show that the proposed novel classification method has better accuracy and efficiency than the existing methods for multiclass problems and has ability to represent the nonlinearity of data structure using a smaller percentage of available data sample.
- 21
- 21
- 21
- Backbone Cannot Be Trained at Once: Rolling Back to Pre-Trained Network for Person Re-Identification20
This work proposes a novel fine-tuning strategy that allows low- level layers to be sufficiently trained by rolling back the weights of high-level layers to their initial pre-trained weights, thereby increasing the performance of ReID tasks.
- 20
- Spatio-temporal weighting in local patches for direct estimation of camera motion in video stabilization20
This paper develops a formula to determine the spatio-temporal weights considering the age, edges, saliency, and distribution information of local patches and verifies that the proposed method gives better performance in camera motion estimation and stabilization of jittering video sequences.
- 20
- 20
- 20
- 19
- 18
- Fast Support Vector Data Description Using K-Means Clustering18
A new fast SVDD method using K-means clustering method that trains each decomposed sub-problems to get support vectors and retrains with the support vectors to find a global data description of a whole target class.
- 17
- 17
- 16
- 16
- 16
- Class-Attentive Diffusion Network for Semi-Supervised Classification15
Adaptive aggregation with Class-Attentive Diffusion (AdaCAD) is proposed, a new aggregation scheme that adaptively aggregates nodes probably of the same class among K-hop neighbors and significantly outperforms the state-of-the-art methods.
- Differentiable Forward and Backward Fixed-Point Iteration Layers14
Experiments show that the fixed-point iteration (FPI) layer can be successfully applied to real-world problems such as image denoising, optical flow, and multi-label classification.
- A Unified Analysis of Switching Multiple Model Adaptive Control - Virtual Equivalent System Approach14
- Construction of neural networks to approximate arbitrary continuous functions of one variable14
It is demonstrated, by examples, that this method yields an easy and efficient construction of neural networks without needing training to approximate a function of one variable.
- 12
- Pose transforming network: Learning to disentangle human posture in variational auto-encoded latent space12
A novel deep conditional generative model to disentangle human posture semantics from image identity in variational auto-encoded latent space is proposed and validated experimentally by demonstrating its pose-transforming capability, outperforming the existing conditionalGenerative model.
- An Integrated Framework for 24-hours Fire Detection12
The integrated framework for 24-hours fire detection with a camera is proposed and the three detectors and the integrated framework are tested with numerous videos, which validates the generality and the robustness of the proposed framework.
- 12
- 12
- 11
- Approximations of lead-time distributions in an assemble-to-order system under a base-stock policy11
An assemble-to-order production system in which a product with multiple components is produced under a base-stock policy and the demand follows a Poisson arrival process, and the production time at each facility is exponentially distributed is considered.
- 10
- Congestion-Aware Bayesian Loss for Crowd Counting10
The estimated person-scale and crowd-sparsity are utilized in the novel congestion-aware Bayesian loss method to improve the supervising representation of the point annotations and achieved state-of-the-art performance.
- Detecting and Removing Text in the Wild10
A new loss function for blending two tasks in a new network structure that depicts wild text regions in a soft mask and selectively inpaints them into a sensible background is proposed to achieve remarkable performance in scene text removal.
- Associative Variational Auto-Encoder with Distributed Latent Spaces and Associators10
The proposed structure successfully associates even heterogeneous modality data and easily incorporates the additional modality to the entire network via the associator.
- 10
- Indirect Adaptive Fuzzy Control of Unmanned Aerial Vehicle10
The model process sensitivity is seen to improve the convergence in addition to improving the response of the UAV when applied to the attitude control of a typical UAV.
- SWAG-Net: Semantic Word-Aware Graph Network for Temporal Video Grounding9
This paper proposes a novel framework called Semantic Word-Aware Graph Network (SWAG-Net), which adopts graph-guided semantic word embedding in an end-to-end manner and outperforms state-of-the-art methods on Charades-STA and ActivityNet Captions datasets.
- Efficient Spatio-Temporal Data Association Using Multidimensional Assignment in Multi-Camera Multi-Target Tracking9
A novel multi-target tracking method which jointly solves a data association problem using images from multiple cameras using a new cost function, considering the accuracy in 3D reconstruction, motion smoothness, visibility from cameras, starting/ending at entrance and exit zone, and false positive.
- View invariant action recognition using generalized 4D features9
A method to recognize human actions independently of viewpoints by developing 4D space-time features which can generalize the information from a finite number of views in training phase so as to show a satisfactory performance in arbitrary testing views is proposed.
- SVDD-Based Method for Fast Training of Multi-class Support Vector Classifier9
A new multi-class classification method based on SVDD (Support Vector Domain Description) that needs only one class data to describe each class is proposed that reduces the order of training time as the size of training data increases.
- 9
- Learnable Negative Proposals Using Dual-Signed Cross-Entropy Loss for Weakly Supervised Video Moment Localization8
Learning-based negative proposals which are trained using a dual-signed cross-entropy loss are proposed and a new evaluation metric is introduced to measure how well a negative proposal captures a poorly-generated positive proposal.
- Improving Knowledge Distillation with Supporting Adversarial Samples8
A new perspective based on a decision boundary, which is one of the most important component of a classifier, is provided and the proposed algorithm trains a student classifier based on the adversarial samples supporting the decision boundaries.
- 8
- Audio Bank: A high-level acoustic signal representation for audio event recognition8
Audio Bank, a new high-level representation of audio, is comprised of distinctive audio detectors representing each audio class in frequency-temporal space and Dimensionality of the resulting feature vector is reduced using non-negative matrix factorization preserving its discriminability and rich semantic information.
- PIL-EYE: Integrated System for Sustainable Development of Intelligent Visual Surveillance Algorithms8
- A Comparison Study of Input ESD Protection Schemes Utilizing NMOS, Thyristor, and Diode Devices8
An in-depth comparison on HBM ESD robustness in terms of lattice heating inside protection devices and peak voltages developed across gate oxides in input buffers, based on DC, mixed-mode transient, and AC analyses utilizing a 2-dimensional device simulator is attempted.
- Feature Extraction Using Class-Augmented Principal Component Analysis (CA-PCA)8
This scheme does not use the scatter matrix including inversion and therefore it is free from the problems of LDA originated from this matrix inversion and therefore it is free from the problems of LDA originated from this matrix inversion.
- 8
- Unified optimization framework for localization and tracking of multiple targets with multiple cameras7
A unified optimization formulation is suggested, which solves two coupled problems simultaneously: the spatio-temporal data association problem and the 3D trajectory estimation problem and a new cost function that describes 3D physical properties of each target is designed.
- 7
- 7
- Use of neural networks in iterative learning control systems7
The neural networks are used to estimate the learning gain of an iterative learning law and to store the learned control input profiles for different reference trajectories and the effectiveness of the present methods has been demonstrated by applying them to a two-link robot manipulator.
- 6
- Motion interaction field for detection of abnormal interactions6
The effectiveness of the proposed method is validated through quantitative and qualitative evaluations with three datasets containing typical abnormal interactions such as car accidents, crowd riots, and uncontrolled fighting.
- Visual tracking of non-rigid objects with partial occlusion through elastic structure of local patches and hierarchical diffusion6
A tracking method based on sequential Bayesian inference that outperforms state-of-the-art on a desktop personal computer (PC) and generates samples from the region where the posterior is concentrated to reduce computation time is proposed.
- Robust pan-tilt-zoom tracking via optimization combining motion features and appearance correlations6
- 6
- Visual tracking with dual modeling6
The proposed method aims to solve the problems of occlusions, background clutters, and drifting simultaneously with the proposed dual model, which is consisted of single Gaussian models for the foreground and the background.
- 6
- 6
- Position-aware Location Regression Network for Temporal Video Grounding5
Position-aware Location Regression Network (PLRN) is proposed which exploits position-aware features of a query and a video and encodes both the video and query using positional information of words and video segments.
- Layer-wise Pruning and Auto-tuning of Layer-wise Learning Rates in Fine-tuning of Deep Networks5
This paper discusses that trends of layer-wise weight variations by fine-tuning using a single learning rate do not match the well-known notion that lower- level layers extract general features and higher-level layers extract specific features, and proposes an algorithm that improves fine- Tuning performance and reduces network complexity through layer- wise pruning and auto-tuned oflayer-wise learning rates.
- 5
- Visual tracking in complex scenes through pixel-wise tri-modeling5
A pixel-wise visual tracking method using a novel tri-model representation, which each model learns the target object, the background, and other non-target moving objects online and uses them in the learning process to prevent drifting.
- Novel Incremental Principal Component Analysis with Improved Performance5
The major contribution of this paper is to explicitly deal with the changing mean and to use a Gram-Schmidt Orthogonalization (GSO) for enforcing the orthogonality of the eigenvectors.
- Perfect Tracking Control for Linear Systems with State Constraint5
A new Perfect Tracking Control scheme for linear systems with state constraint is presented, which increases the number of the steps on-line for perfect tracking to satisfy the given ellipsoid-type state constraint.
- 5
- 5
- Development of Reconfigurable and Evolvable Architecture for Intelligence Implement5
This paper proposes an Evolvable and Reconfigurable ERI Architecture based on intelligent Macro Core (IMC) so that various and new algorithms can be easily added incrementally and construct the reconfigured intelligent system easily.
- 5
- 5
- How Much a Model be Trained by Passive Learning Before Active Learning?4
This paper proposes a method to automatically determine the starting point of active learning for the given settings and develops three empirical formulas to determine an appropriate entropy of sample-uncertainty that should be obtained by passive learning before starting active learning.
- Rotation-Aware 3D Vehicle Detection From Point Cloud4
This paper proposes a rotation-aware 3D vehicle detector that extracts distinguishable features from the proposals with various angles of rotations and improves detection accuracy as well as rotation angle estimation accuracy against the existing algorithms without much loss of computational efficiency.
- 4
- Combining histogram-wise and pixel-wise matchings for kernel tracking through constrained optimization4
This paper proposes a constrained optimization approach to improving both the robustness and accuracy of kernel tracking which is appropriate for real-time video surveillance due to its low computational load and outperforms recent tracking algorithms in terms of robustness, accuracy, and computational cost.
- 4
- 4
- Detection of suspicious activity using incremental outlier detection algorithms4
Efficiency of various techniques for incremental outlier detection on detecting unusual trajectories on simulated and real-life data obtained from SENSIAC database are compared.
- 4
- 4
- 4
- TextGuider: Training-Free Guidance for Text Rendering via Attention Alignment3
This work proposes TextGuider, a novel training-free method that encourages accurate and complete text appearance by aligning textual content tokens and text regions in the image by analyzing attention patterns in Multi-Modal Diffusion Transformer (MM-DiT) models.
- Motion-aware ensemble of three-mode trackers for unmanned aerial vehicles3
The proposed motion-aware ensemble shows a 5.3% improvement in average precision compared to the baseline correlation filter tracker, which effectively employs deep features while achieving a tracking speed of at least 80fps in the authors' experimental settings.
- Re-ranking with ranking-reflected similarity for person re-identification3
A novel re-ranking method using a ranking-reflected metric to measure the similarity between the ordered set of K-nearest neighbors of a probe and that of a gallery and shows that the proposed method improves the Re-ID accuracy by add-on to the state-of-the-art methods.
- 3
- 3
- Density-Aware Pedestrian Proposal Networks for Robust People Detection in Crowded Scenes3
Experimental results show that the DAPPN method outperforms the conventional method and achieves near real-time speed on a GPU (25 fps).
- Graph pyramids for protein function prediction3
The Protein-Protein Similarity (PPS) network is constructed, which captures the subtle properties of protein families by examining the interactions among 'weakly interacting proteins' in the PPS network and by using hierarchical graph analysis via the graph pyramid.
- Robust Feature Extraction for Shift and Direction Invariant Action Recognition3
Experiments show that the proposed feature based on optical flow for action recognition gives a good performance comparable to the state-of-the-art approaches and has applicability to actual environments with smart class dataset.
- Learning with Adaptive Rate for Online Detection of Unusual Appearance3
Experimental results show the effectiveness of the proposed method to detect unusual appearances qualitatively and quantitatively to avoid the biased learning problem and reduce the noisy detection.
- A Thyristor-Only Input ESD Protection Scheme for CMOS RF ICs3
An input protection scheme composed of thyristor devices only without using a clamp NMOS device is proposed in order to minimize the area consumed by a pad structure in CMOS RF ICs and a detailed analysis on the problems which can occur in real HBM tests is attempted.
- 3
- Face feature processor on mobile service robot3
This research considered face detection, face tracking and face recognition from continuous camera image, which were implemented on mobile service robot OMR to verify.
- 3
- Adaptive nonlinear equalizer with reduced computational complexity3
Simulation results show that the proposed nonlinear equalizer can be effectively applied to on-line adaptation environments due to its reduced computational complexity, making complexity similar to that of a linear equalizer.
- 2
- 2
- 2
- Weighted Pooling Based on Visual Saliency for Image Classification2
A novel pooling method which decomposes an image into an object region and a background region, and adopts the visual saliency map for a probabilistic measure of objectness to estimate object region from an unlabeled image.
- 2
- 2
- Robust Salient Moving Object Detection with Light-Computational Load2
A method which detects salient moving objects with light-computational load by using background modeling and subtraction method which is similar to adaptive threshold with foreground map and generated by Modified Temporal Difference to speed up the process.
- 2
- Robust Motor Speed Control under Time Varying Loads in Moving Actuator Type artificial Heart (AnyHeart)2
The proposed control technique overcomes the limitations of the PI controller, and makes important improvements in both performance and stability.
- C2I-CAT: Class-to-Image Cross Attention Transformer for Out-of-Distribution Detection1
The proposed cross-attention structure with score ensemble inference improves largely near OOD detection performance, where FPR95 improvement in near OOD detection compared to the state-of-the-art method becomes 2.55% for CIFAR-10 and 2.67% for CIFAR-100, keeping competitive classification accuracy.
- 1
- 1
- Differentiable Fixed-Point Iteration Layer1
It is shown that the derivative of an FPI layer depends only on the fixed point, and then a method to calculate it efficiently using another FPI which is called the backward FPI is presented.
- 1
- 1
- Proximity Clustering for Revealing a Semantically Dominant Class1
The proximity clustering (PxC) algorithm, which finds the spatially diverse but semantically similar data-point colonies in the form of a dominant class cluster, improves performance of the image classification and the unsupervised image segmentation.
- 1
- 1
- Action Chart: A Representation for Efficient Recognition of Complex Activity1
Experimental results show that the Action Chart gives promising recognition performance with low computational load and can be used for abstracting long video sequences.
- Hierarchical feature grouping for multiple object segmentation and tracking1
A hierarchical feature grouping method for multiple object segmentation and tracking that efficiently segments and tracks various kinds of objects without object detector, 3D model and geometry information.
- An Analytical Approach for Fast Automatic Sizing of Narrow-Band RF CMOS LNAs1
A fast automatic sizing algorithm for a single-ended narrow-band CMOS cascode LNA adopting an inductive source degeneration based on an analytical approach without any optimization procedure is introduced, which shows one way how accurate automatic synthesis can be done in a analytical approach.
- 1
- Orientation and Scale Invariant Kernel-Based Object Tracking with Probabilistic Emphasizing1
The proposed new mean-shift based tracking algorithm is shown to be robust to background clutter and is able to track complex movements very accurately even in shaky scenarios.
- 1
- 1
- 1
- 1
- 1
- Neural-Based Iterative Learning Control1
This chapter presents a neural-based iterative learning control method for unknown nonlinear systems used for a process of short term memory involved in a temporary learning manipulation and a storage of a specific temporary action.
- Robust adaptive control of feedback linearised systemswith time-varying disturbances1
A robust adaptive control scheme is proposed to reject the unknown bounded time-varying disturbances in the feedback linearised control system and the asymptotic output tracking is proved by using the Lyapunov stability theory.
- Computationally Efficient Fuzzy Adaptive Equalizer for Nonlinear Channels1
The proposed equalizer shows a good performance with a computational load similar to that of linear equalizers at each symbol time and the effectiveness of the approach is shown through computer simulations of bit error rate performances and computational complexity.
- 1
- 1
- –
- Consensus and Risk Aversion Learning in Ensemble of Multiple Experts for Long-Tailed Classification–
A new learning method for expert ensemble is proposed, which obtains the consensus by aggregating the predictions of experts and maximizes the expected prediction accuracy of each expert without excessive diversity from the consensus.
- –
- –
- –
- –
- –
- Heterogeneous Double-Head Ensemble for Deep Metric Learning–
This paper proposes a heterogeneous double-head ensemble structure that drastically increases ensemble gain along with memory savings and outperforms the state-of-the-art algorithms in verifying experiments on image retrieval datasets.
- –
- A Comparison Study of Diode-String ESD Clamps for CMOS Input Protection–
It is shown that the diode-string devices in comparison can serve successfully as a VDD-VSS clamp device for ESD protection by virtue of the dominant pnpn thyristor-related conduction mechanisms.
- –
- –
- –
- –
- Erratum to: Visual tracking in complex scenes through pixel-wise tri-modeling–
This poster presents a probabilistic simulation of the response of the immune system to EMTs to treat central nervous system-derived central giant cell granuloma.
- Traffic Pattern Analysis and Anomaly Detection via Probabilistic Inference Model–
This chapter introduces a method for trajectory pattern analysis through the probabilistic inference model with both regional and velocity observations, and suggests a novel learning scheme instead of collapsed Gibbs sampling to achieve online learning even with the complexity of the proposed model.
- Multiple ground plane estimation for 3D scene understanding using a monocular camera–
This paper proposes a method for estimation of multiple ground planes using a stationary monocular camera and performs back-projection with the votes from the first step to extract trajectories which support each ground plane.
- Linear boundary discriminant analysis based on QR decomposition–
This work reformulate LBDA using QR decomposition, and this results in both reducing computation time and resolving SSS problem while classification performance is maintained.
- –
- –
- Adaptive Visual Tracking for Surveillance Systems–
A new hierarchical estimation approach for adaptive visual tracking that includes incremental appearance model and hierarchical estimation: global estimation and local estimation, which enables real-time tracking while maintaining or improving tracking abilities.
- Speeded-up Algorithm for Human/Vehicle Classification using Hilbert Scanning Distance–
This algorithm is similar to previous pixel-based template matching scheme like Hausdorff Distance, but it uses 1D image array rather than 2D regions inspired by Hilbert Curve, and shows robustness and real-time performance in human/vehicle classification.
- Block Type Modular Robot: Mom's Friend–
This paper introduced a block type modular service robot system, “Mom's friend”, which implemented eight block modules for building block type reconfigurable intelligent modular robot and assembled various types of robot using these block modules.
- Support Vector Data Description Using Clustering Method–
This work proposes SVDD using clustering method, similar to divideand- conquer strategy; trains each decomposed subproblems to get support vectors, and uses two different clustering algorithms; K-means clustering and Mean Shift clustering.
- System Architecture of Store Service Robot–
This paper proposes system architecture for intelligent service robot that several intelligent functions are modulaized so that it can work individually and will help to develop robot easily.
- A novel algorithm for polyp detection using Eigen decomposition of Hessian-matrix for CT colonography CAD: validation with physical phantom study–
The algorithm is clinically applicable for polyp detection, because of high sensitivity and relatively low false positive detections, and Gaussian blurring factor and shape threshold parameters are optimized.
- –
- –
- PCA-FX : A New Supervised Feature Extraction Scheme based on PCA–
This paper proposes a feature extraction scheme which uses class information to extract features by PCA and test the algorithm using Yale face database, and analyzes the performance to compare with other algorithms.
- SVM classifier using Features Extracted by PCA For face recognition–
This paper proposes multi-class support vector machines as a appropriate classifier for feature extraction method by PCA for face recognition and shows that proposed scheme has better performance than other schemes and is adequate for PCA using class information.
- –
- –
- –
- –
- –
- An output feedback tracking controller for systems diffeomorphic to the non-linear observer form–
An output feedback tracking controller is proposed for single-input, single-output non-linear systems that are diffeomorphic to the non- linear observer form and the backstepping control method is adopted based on the state estimates.
- –
- –
- –
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.