Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks189 from public data
- 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.
- 434
- Content based image retrieval using motif cooccurrence matrix308
The retrieval using the MCM is better than the CCM since it captures the third order image statistics in the local neighborhood and the use of MCM considerably improves the retrieval performance.
- 231
- Median Binary Pattern for Textures Classification158
It is demonstrated - using a set of classifications, that the proposed method significantly improves the capability of texture recognition and outperforms classical algorithms.
- 155
- 125
- 102
- 85
- VisDrone-VDT2018: The Vision Meets Drone Video Detection and Tracking Challenge Results81
A large-scale video object detection and tracking dataset, which consists of 79 video clips with about 1.5 million annotated bounding boxes in 33, 366 frames, and the evaluation protocol of the VisDrone-VDT2018 challenge and the results of the algorithms on the benchmark dataset are presented.
- 80
- 75
- Wide-Area Persistent Airborne Video: Architecture and Challenges70
Traditional tradeoffs posed between lens size and resolution, that is the numerical aperture of the system, can now be mitigated using an array of cameras mounted in a specific geometry, enabling new imaging systems to cover very large fields of view at high resolution, albeit with spatially varying point spread functions.
- 68
- Joint Adaptive Median Binary Patterns for texture classification68
The proposed novel texture descriptor, the Adaptive Median Binary Pattern (AMBP), performs better than the best state-of-the-art techniques in the noiseless case and significantly outperforms all of them in the presence of impulse noise.
- The Thermal Infrared Visual Object Tracking VOT-TIR2016 Challenge Results66
A significant general improvement of results has been observed, which partly compensate for the more difficult sequences introduced into the dataset in VOT-TIR2016.
- Application Of Papoulis-Gerchberg Method In Image Super-Resolution and Inpainting64
This paper studies the Papoulis–Gerchberg (PG) method and its applications to domains of image restoration such as super-resolution (SR) and inpainting and suggests improvements to the method to achieve better SR and inpainted results.
- 59
- 52
- 49
- Rotationally Invariant Hashing of Median Binary Patterns for Texture Classification48
A novel image feature descriptor for rotationally invariant 2D texture classification that outperforms the rotation invariant uniform local binary patterns (LBP UNIF) and quantitatively compare and demonstrate the advantage of the new MBP texture descriptors for classification using the Brodatz and Outex texture dictionaries.
- 47
- Summary of methods in Wide-Area Motion Imagery (WAMI)44
This paper summarizes the published literature on WAMI results in an effort to organize the techniques, discuss the developments, and determine the state-of-the-art.
- 44
- Dynamic Data Driven Applications System Concept for Information Fusion44
A Wide-Area Motion Imagery application is used to draw parallels and contrasts between IF and DDDAS systems that warrants an integrated perspective and is aimed at triggering a sequence of deeper insightful research towards exploiting sparsely sampled piecewise dense WAMI measurements.
- 43
- Dynamic Data Driven Applications Systems (DDDAS) modeling for automatic target recognition42
This elementary work uses a Wide-Area Motion Imagery application to draw parallels and contrasts between ATR and DDDAS systems that warrants an integrated perspective and triggers a sequence of deeper insightful research towards exploiting sparsely sampled piecewise dense WAMI measurements.
- 41
- 40
- 39
- 39
- 38
- 36
- 35
- 33
- Overview of contextual tracking approaches in information fusion32
Five contextual information categories that support target tracking are described that include domain knowledge from a user to aid the information fusion process through selection, cueing, and analysis, environment-to-hardware processing for sensor management, known distribution of entities for situation/threat assessment, historical traffic behavior for situation awareness patterns of life (POL), and road information for target tracking and identification.
- 30
- Pattern Activity Clustering and Evaluation (PACE)30
This work seeks to explore contemporary methods in activity analysis using machine learning techniques that discover and characterize behaviors that enable grouping, anomaly detection, and adversarial intent prediction.
- 29
- VisDrone-MOT2020: The Vision Meets Drone Multiple Object Tracking Challenge Results27
The VisDrone-MOT2020 consists of 79 challenging video sequences, including 56 videos for training, 7 videos for validation and 17 videos for evaluation, which feature 12 participating MOT algorithms.
- 27
- Heterogeneous Multi-View Information Fusion: Review of 3-D Reconstruction Methods and a New Registration with Uncertainty Modeling27
A framework to incorporate measurement uncertainties in the registered imagery is proposed, which is a critical issue to ensure the robustness of these applications but is often not addressed.
- 26
- 26
- 26
- 25
- ShareBoost: Boosting for Multi-view Learning with Performance Guarantees25
This paper proposes an algorithm based on boosting that builds base classifiers independently from each data type (view) that provides a partial view about an object of interest and introduces a randomized version of the algorithm, where a winning view is chosen probabilistically.
- Parallel Blob Extraction Using the Multi-core Cell Processor25
Novel parallelization and explicit instruction level optimization techniques are described for fully exploiting the computational capacity of the Synergistic Processing Elements (SPEs) on the Cell processor.
- 24
- 23
- Cognitive Computing Systems: Their Potential and the Future22
The articles in this special section focus on cognitive computing systems, autonomous systems that are based on machine cognition, which refers to processes and systems that enable humans to perform both mundane and specialized tasks.
- 22
- 20
- 20
- Parallel Implementation of the Integral Histogram20
A general parallel version of the the integral histogram is formulated and its implementation in Star Superscalar (StarSs) is analysed to provide extensive performance measurements and tradeoffs using two different scan orders or histogram propagation methods.
- 20
- VISUAL TRACKING WITH ROBUST TARGET LOCALIZATION19
A tracking method with robust target localization for tracking of visual objects using an adaptive appearance model that incorporates structural information to avoid drifts and can be updated incrementally using partial models is presented.
- Functionally important segments in proteins dissected using Gene Ontology and geometric clustering of peptide fragments19
Clustering-cum-statistical analysis for peptide fragments 8 residues in length and with only trans peptide bonds shows that molecular function propensities ≥20 and p-values ≤0.05 can dissect fragments within a protein linked to the molecular function.
- 18
- 18
- 18
- An approach to interactive retrieval in face image databases based on semantic attributes17
Personal Construct Theory (PCT) is introduced as a knowledge elicitation tool for systematically deriving semantic attributes to support RSA in image retrieval applications and an algorithm for RSA based on the repertory grid is proposed.
- Detection of potential mosquito breeding sites based on community sourced geotagged images16
This work shows how the geotagged images collected through the general population can be used to combat Malaria and Dengue by identifying and visualizing localities that contain potential mosquito breeding sites.
- Efficient GPU Implementation of the Integral Histogram16
A proposed Graphics Processing Unit (GPU) implementation uses parallel prefix sums on row and column histograms in a cross-weave scan with high GPU utilization and communication-aware data transfer between CPU and GPU memories.
- 16
- Content Based Image Retrieval Using Motif Cooccurence Matrix.16
Experiments show that use of MCM considerably improves the performance of CBIR system and performs much better than CCM since it captures the third order image statistics in the local neighborhood.
- DeepFTSG: Multi-stream Asymmetric USE-Net Trellis Encoders with Shared Decoder Feature Fusion Architecture for Video Motion Segmentation15
A novel deep architecture for robust moving object detection that incorporates single and multi-stream multi-channel USE-Net trellis asymmetric encoders extending U-Net with squeeze and excitation (SE) blocks and a single shared decoder network for fusing multiple motion and appearance cues is proposed.
- 15
- Feature selection for appearance-based vehicle tracking in geospatial video15
Overall SFFS and SFS perform very well, close to the optimum determined by FOCUS, but RELIEF does not work as well for feature selection in the context of appearance-based object tracking.
- Interactive target tracking for persistent wide-area surveillance15
This work developed an algorithm that overcomes challenges in tracking targets in persistent aerial surveillance and demonstrated it on data obtained from a wide-area imaging platform.
- 15
- 15
- 14
- Calibration of camera parameters using vanishing points14
A new technique for calibrating the intrinsic and extrinsic parameters of a video imaging system, based on vanishing points is presented, which incorporates the radial distortion factor and the basic equations become highly nonlinear (eighth order).
- CajunBot: Architecture and algorithms13
CajunBot, an autonomous ground vehicle and a finalist in the 2005 DARPA Grand Challenge, is built on the chassis of MAX IV, a six‐wheeled all‐terrain vehicle (ATV), which required adding drive‐by‐wire control, lidar sensors, an inertial navigation system (INS), and a computing system.
- A NOVEL REPRESENTATION AND ALGORITHMS FOR (QUASI) STABLE MARRIAGES13
This paper defines 3 algorithms corresponding to 3 different scans of the marriage table to meet progressively all constraints and proposes ”stable marriages” algorithms based on a novel representation called marriage table.
- Velocity vectors for features of sequential oceanographic images13
An approximation method to characterize the deformation of these image motions for the purpose of estimating the velocity field of these images, focused on the interpretation of the change in the extracted features of the Gulf Stream.
- GLSNet++: Global and Local-Stream Feature Fusion for LiDAR Point Cloud Semantic Segmentation Using GNN Demixing Block12
This work developed global and local stream deep network (GLSNet++), an innovative deep learning architecture for robust context-dependent 3-D point cloud segmentation that can be extended to more classes, higher-resolution point clouds, and larger geographic regions.
- Stabilization of Airborne Video Using Sensor Exterior Orientation with Analytical Homography Modeling12
An analytic image stabilization approach is described in this chapter where pixel information from the focal plane of the camera is stabilized and georegistered in a global reference frame to derive a direct closed-form analytic expression from 3D camera poses that is robust even in the presence of significant scene parallax.
- Geospatial content summarization of UAV aerial imagery using mosaicking12
A feature based registration method is proposed which brings the complexity and homogeneity nature of maize data under consideration and formulates a two step registration method which reduces error accumulation acquired through traditional frame-to-frame registration.
- 3D patch-based multi-view stereo for high-resolution imagery12
An improved solution to image-based three-dimensional (3D) modeling (also known as ”multi-view stereo”) that outputs surfaces visible in high-resolution wide-area format video also known as widearea motion imagery (WAMI) consisting of a dense set of small 3D points.
- Dismount tracking and identification from electro-optical imagery12
The aim is to provide a discussion to ensure that nominated combatants are detected, the sensed information is validated across multiple perspectives, the reported confidence values achieve positive combatant versus non- combatant detection, and the related situational awareness attributes are provided in a timely and reliable manner to stakeholders.
- 12
- 12
- 11
- 11
- Feature fusion using ranking for object tracking in aerial imagery11
A framework for dynamic online ranking-based feature evaluation and fusion in aerial wide-area tracking that increases localization accuracy by reducing multimodal features or background clutter and increases the robustness of the tracker in dynamically changing environments.
- 10
- Multichannel texture image segmentation using local feature fitting based variational active contours10
The novel contributions include the use of local feature density functions in the context of a luminance-chromaticity decomposition combined with a globally convex active contour variational method to capture texture variations for image segmentation.
- 9
- 9
- 8
- 8
- 8
- 8
- 8
- KOLAM: a cross-platform architecture for scalable visualization and tracking in wide-area imagery8
This paper focuses on the use of KOLAM for target tracking in high-resolution, high throughput wide format video also known as wide-area motion imagery (WAMI), and develops a supervised tracking and visualization tool that allows analysts to digitally tag multiple targets, quickly review and correct tracking results and apply geospatial visual analytic tools on the generated trajectories.
- 8
- 7
- 7
- 6
- Realtime motion detection based on the spatio-temporal median filter using GPU integral histograms6
This work describes a parallel 3D spatiotemporal median filter algorithm implemented in CUDA for many core Graphics Processing Unit (GPU) architectures using the integral histogram as a building block to support adaptive window sizes and characterize performance of the parallel 3d median filter for different image sizes and varying number of histogram bins.
- Feature prominence-based weighting scheme for video tracking6
Experimental results show that feature prominence outperforms these other feature fusion methods and is embedded within a test-bed tracking system.
- Information expectation from unmanned aircraft swarms6
This paper lays the groundwork for modeling the quantification of sensor coverage for Unmanned Aircraft (UA) swarms of sensors by experimentation using simulation, and gains insight into the quantifiable influence of varying swarm sizes and configurations on area coverage.
- Next Generation Distributed Sensor Networks6
Some of the exciting information processing problems that are being solved to effectively harvest the benefits of current and emerging nano, micro, and macro sensors in distributed sensor networks are introduced.
- 5
- 5
- The effect of state dependent probability of detection in multitarget tracking applications5
Using a simulation environment inspired by the challenges of Wide Area Surveillance, a state dependent formulation for probability of detection and clutter is developed and is compared in a simulated urban environment populated by multiple vehicles and cursed with occlusions.
- Exploiting fisher and fukunaga-koontz transforms in chernoff dimensionality reduction5
It is shown that the Chernoff criterion exploits the Fisher and Fukunaga-Koontz transforms in computing its linear discriminants and that a recently proposed decomposition of the data space into four subspaces is incomplete.
- 5
- 5
- 5
- Video-Assisted Global Positioning in Terrain Navigation with Known Landmarks5
A robust Hough transform-like method, facilitated by a class of CORDIC-structured computations is developed to find the camera position followed by a method of computing the position of a ground object from images of that object and three known landmarks, enabling fast and effective visual terrain navigation of aerial surveillance systems when the global positioning and inertial navigation sensors become faulty, inaccurate, or dysfunctional.
- 5
- 5
- 4
- 4
- 4
- Evaluation of feature matching in aerial imagery for structure-from motion and bundle adjustment4
Experimental results demonstrate that SURF detector combined with SURF descriptor generates the longest feature tracks while FAST detector plus SIFT descriptor produces the highest matching precision.
- Real-time geoprojection and stabilization on airborne GPU-enabled embedded systems4
The challenges and capabilities of implementing a fast, efficient georegistration system on low-power embedded systems are described and future application in real-time on-device processing, given initial speeds on embedded systems is proposed.
- 4
- 4
- 4
- 4
- Wide-area video exploitation (WAVE) joint data management (JDM) for layered sensing4
The various issues surrounding Wide-Area Video Exploitation (WAVE) in a layered-sensing environment are explored to include improvements in Joint Data Management such as data collection, construction, and transformation, and information evaluation, presentation, and dissemination.
- Multiscale moment transforms over the integer lattice4
After introducing a mathematically well-behaved method for rotating an image over the two-dimensional integer lattice, it is shown that this rotation commutes with the moment-based transform in the expected manner.
- 4
- 4
- A simplified design strategy for mapping image processing algorithms on a SIMD torus4
This feature facilitates coordinated use of several independently operating SIMD meshes in a newly emerging computer vision paradigm known as multiview image-sequence analysis (MVISA) for 3-D perception of unstructured dynamic scenes.
- 3
- Blind restoration of aerial imagery degraded by spatially varying motion blur3
This paper develops a methodology for blind restoration of spatially varying blur induced by camera motion caused by instabilities of the moving platform and proposes a scheme to estimate the original focused image affected by arbitrarily-shaped blur kernels.
- 3
- 3
- Chernoff Dimensionality Reduction–Where Fisher Meets FKT3
What can be expected from the Chernoff criterion and its relations to the Fisher criterion and Fukunaga-Koontz transform are shown and it is shown that a recently proposed decomposition of the data space into four subspaces is incomplete.
- 3
- 3
- 2
- Short-Term Video Stabilization Using Ground Plane Segmentation for Low Altitude UAV Object Tracking (Conference Presentation)2
This work proposes a robust video stabilization algorithm which allows the flux motion detection to efficiently identify moving objects and shows satisfactory results when other methods shown to fail on the same type of raw videos.
- Csang: continuous scale anisotropic Gaussians for robust linear structure extraction2
An adaptive robust structure Tensor (ARST) method where the local adaptation process uses spatially varying adaptive Gaussian kernel that is initialized using the total least-squares structure tensor solution to obtain robust edge and junction features.
- Illumination robust change detection with CMOS imaging sensors2
An optimization problem where each row of the distorted image is approximated as a weighted sum of the corresponding rows in warped versions of the reference image due to camera motion within the exposure period to account for geometric as well as photometric differences is formulated.
- 2
- 2
- Multisensor Network‐Based Framework for Video Surveillance: Real‐Time Superresolution Imaging2
A network of multi modal sensors with distributed and embedded computations is proposed for a video surveillance and monitoring application and the overall performance is promising when compared against that of a small network of cameras monitoring large corridors with a human operator in the loop.
- 2
- 2
- Fundamental Constraints on Camera Centers Arising from 3D-2D Matches of Points and Lines1
The minimal cases for which n point matches and m line matches constrain the camera center to lie on an algebraic surface in R3 are addressed.
- Optimizing short-term data association for multiple object tracking1
The LSAP algorithm was modified to use adjacency lists to optimize the mapping process instead of relying on a full cost matrix in the MJV algorithm and incorporated optimized data structures like min-heaps and unordered sequential lists, which significantly improved performance on geospatial aerial drone tracking datasets.
- 1
- pyTAG: python-based interactive training data generation for visual tracking algorithms (Conference Presentation)1
The proposed tool's plugin structure allows integration, testing, and validation of different trackers, and the tracker can be paused, resumed, forwarded, rewound and re-initialized on the run, after it loses the object.
- EpiX: A 3D Measurement Tool for Heritage, Archeology, and Aerial Photogrammetry1
This paper presents an open, cross-platform, effective and extensible GUI annotation tool named EpiX, exploiting the geometric features of epipolar lines, for large photogrammetric imagery analysis, applicable to acquire airborne images of archeological and heritage sites.
- Multi-Scale Spatially Weighted Local Histograms in O(1)1
A novel fast algorithm to evaluate spatially weighted local histograms at different scale accurately and in constant time using an extension of integral histogram to provide an efficient exhaustive search, improve target localization accuracy and meet the demand of real-time processing.
- 1
- Camouflaging Motion Blur1
This work applies a suitably derived blurring model to smear a target object and naturally blend it into the motion blurred background by using alpha matting to raise the interesting issue of camouaging motion blur with special relevance to non-uniformly blurred images.
- 1
- 1
- ACM SRC poster: a portable implementation of the integral histogram in starss1
IH in StarSs is the first parallel implementation of this algorithm in the literature, and the applications for StarPU, MAGMA and PLASMA, which so far have been limited to numerical linear algebra kernels, are targeted.
- 1
- 1
- 1
- 1
- 1
- Video-Assisted Global Positioning in Terrain Navigation: Hough Transform Solution.1
A robust Hough transform-like method, facilitated by a class of CORDIC-structured computations, is shown feasible within the framework of terrain navigation, and empowers aerial surveillance systems to navigateively when the global position and inertial navigation sensors become faulty, inaccurate, or dysfunctional.
- –
- –
- –
- –
- –
- A hybrid local and global multi-object tracking with semantic spatial and appearance modules–
A hybrid tracking-by-detection system that combines local and global data association scheme to ensure efficiency and reduce complexity and can handle the appearance of new objects, temporal disappearance, object terminations, and object occlusions.
- –
- Point based interactive image segmentation using multiquadrics splines–
Quantitative and qualitative results on the standard datasets show that MQ outperforms other regression based methods, GEBS, Ridge Regression and Logistic Regression, and popular methods like Graph Cut,4 Random Walk and Random Forest.
- –
- Cueing motion blur for registration of inclined planar scenes–
This work extends the state-of-the-art by cueing motion blur itself to infer plane inclination by matching extremities of blur kernels computed at different locations in the image by showing that it is possible to find the orientation.
- –
- –
- Intelligent Autonomous Systems: A Layered Architecture in the Age of Multicore Processing–
The talk will highlight the evolution of signal processing, linguistic hierarchy, computing models where information fusion is considered, and draw some parallels to make a case for a layered architecture for context-adaptive autonomous systems to swiftly change the behaviors of intelligent systems.
- Margin Based Dimensionality Reduction and Generalization~!2010-03-17~!2010-05-26~!2010-08-23~!–
This paper proposes a margin based criterion for linear dimension reduction that addresses the above problem associated with LDA and establishes an error bound for the proposed technique by showing its relation to least squares regression.
- –
- –
- –
- –
- –
- –
- –
- Outlier accommodation using genetic algorithms in neural networks-based image segmentation–
The research presented in this report presents a neural network approach to find an optimal-edge detector able to segment a wide variety of images, and a more robust and unsupervised search technique through a genetic algorithm.
- –
- –
- –
- –
- –
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.