Abstract Stereo videography is a powerful technique for quantifying the kinematics and behavior of animals, but it can be challenging to use in an outdoor field setting. We here present workflow associated software performing calibration cameras placed setting estimating accuracy resulting stereoscopic reconstructions. demonstrate through example reconstructions bat bird flight. provide tools planning experiments processing calibrations that other researchers may calibrate their own cameras....

We hereby publish a new thermal infrared video benchmark, called TIV, for various visual analysis tasks, which include single object tracking in clutter, multi-object or multiple views, analyzing motion patterns of large groups, and censusing wild animals flight. Our data describe real world scenarios, such as bats emerging from their caves numbers, crowded street view during marathon competition, students walking through an atrium class break. also introduce baseline methods evaluation...

We present a novel framework for multiple object tracking in which the problems of detection and data association are expressed by single objective function. The follows Lagrange dual decomposition strategy, taking advantage often complementary nature two subproblems. Our coupling formulation avoids problem error propagation from traditional "detection-tracking approaches" to suffer. also eschew common heuristics such as "nonmaximum suppression" hypotheses modeling joint image likelihood...

We propose a randomized ensemble algorithm to model the time-varying appearance of an object for visual tracking. In contrast with previous online methods updating classifier ensembles in tracking-by-detection, weight vector that combines weak classifiers is treated as random variable and posterior distribution estimated Bayesian manner. essence, reflects confidence among used construct adapt ensemble. The resulting formulation models discriminative ability so ensembled strong can varying...

Location positioning in indoor environments is a major challenge. Various algorithms have been developed over years to address the problem of positioning. One most cost effective choice for based on received signal strength indicator (RSSI) using existing Wi-Fi networks commercial and/or public areas. This solution infrastructure-free and offers meter-range accuracy. In this paper, machine learning approaches including k-nearest neighbor (k-NN), rules-based classifier (JRip), random forest...

We propose a multi-object multi-camera framework for tracking large numbers of tightly-spaced objects that rapidly move in three dimensions. formulate the problem finding correspondences across multiple views as multidimensional assignment and use greedy randomized adaptive search procedure to solve this NP-hard efficiently. To account occlusions, we relax one-to-one constraint one measurement corresponds object iteratively relaxed problem. After are established, trajectories estimated by...

Pavement preservation programming often involves multiple objective considerations. Traditional single-objective optimization frameworks consider one single while imposing competing objectives as constraints in the formulation. This paper proposes an approach for pavement that uses multiobjective and chance constraints. can handle incommensurable conflicting considering probabilistic related to available budget over planning horizon. The implementation of a case study with two of: (1)...

This paper proposes novel algorithms that use network-flow and set-cover techniques to perform occlusion reasoning for a large number of small, moving objects in single or multiple views. We designed track-linking framework about short-term long-term occlusions. introduce two-stage process automatically construct "track graph" describes the track merging splitting events caused by occlusion. To explain occlusions, when local information is sufficient distinguish objects, links trajectory...

Optimization-based tools have been included in many engineering management systems for individual infrastructure asset classes such as pavement (PMS) and bridge (BMS). These typically include single-objective optimization analyses. However, real-world decisions concerning preservation renewal often involve more than one objective reflecting the various goals of agency need to evaluate possible alternatives according multiple criteria. Traditional approaches handling situations optimize a...

This paper proposes an indoor localization method using online independent support vector machine (OISVM) classification and undersampling techniques. The system is based on the received signal strength indicator (RSSI) of Wi-Fi signals. A new algorithm developed to address imbalanced data problem associated with OISVM, a kernel function parameter selection introduced for training process. time complexity both process prediction are significantly decreased. Comparative experimental results...

This paper compares different pavement performance models (PPMs) from around the world and recommends one for use in Portuguese management systems (PMS). The is divided into three complementary sections. first section describes several PPMs used world. analysed are (or come from) Highway Development Management System, American Association of State Transportation Officials, Nevada PMS, Collop–Cebon whole-life performance, Swedish PMS Spanish PMS. second results obtained a case study with...

In the context of rapid advancements in autonomous driving technology, ensuring passengers' safety and comfort has become a priority. Obstacle or road detection systems, especially accurate pavement condition identification unfavorable weather time circumstances, play crucial role safe operation comfortable riding experience vehicles. To this end, we propose novel framework based on image quality enhancement feature distillation (IQEFD) for detecting diverse conditions during day night to...

Generally, an alcoholic's brain shows explicit damage. However, in cognitive tasks, the correlation between topological structural changes of networks and damage is still unclear. Scalp electrodes synchronization likelihoo

We developed two methods for tracking multiple objects using several camera views. The use the Multiple Hypothesis Tracking (MHT) framework to solve both across-view data association problem (i.e., finding object correspondences across views) and across-time assignment of current measurements previously established tracks). ¿tracking-reconstruction method¿ establishes two-dimensional (2D) tracks each view then reconstructs their three-dimensional (3D) motion trajectories....

In order to solve the problem of building corresponding relationship between attacks and responses scientifically in AIR (Automated Intrusion Response) systems, a taxonomy network computer based on response is proposed this paper. Through analyzing existing researches attack classification special need intrusion response, new constructed by classifying into three main dimensions subdividing them from perspectives response. The first dimension aimed classify localities initiate, second covers...

This paper proposes a fast online multi-target tracking method, called motion agreement algorithm, which dynamically selects stable object regions to track. The appearance of each object, here pedestrians, is represented by multiple local patches. For patch, the algorithm computes estimate direction motion. By fusion agreements between global and estimate, identifies enables robust tracking. proposed patch-based model was integrated into an efficient system that uses bipartite matching for...

Functional connectivity networks (FCN) are the physiological basis of brain synchronization to integrating neural activity. They not rigid but can reorganize under pathological conditions or during mental behavioral states. However, because acts be very fast, like blink an eye, we now used visual system as a model explore rapid FCN reorganization and its functional impact in normal, abnormal post treatment vision. EEG-recordings were time-locked stimulus presentation; graph analysis...

Because of a lack reliable performance prediction models, many state departments transportation (DOTs) use needs-based budgeting process, namely, annual budget requests. Allocations funding across maintenance activities and districts are developed on the basis pavement needs derived from inventory or biannual condition data. This allocation is challenging often involves negotiation balancing. A decision support model proposed for optimization short-term preservation based two proven...

We have developed methods for segmentation and tracking of cells in time-lapse phase-contrast microscopy images. Our multi-object Bayesian algorithm detects tracks large numbers presence clutter identifies cell division. To solve the data association problem, assignment current measurements to tracks, we tested various cost functions with both an optimal a fast, suboptimal algorithm. also propose metrics quantify migration properties, such as motility directional persistence, compared our...

The nature of forest structure plays an important role in the study foraging behaviors bats. In this study, we demonstrate a new combined methodology that uses both thermal imaging technology and ground-based LiDAR system to record reconstruct Eptesicus fuscus (big brown bats) flight trajectories three-dimensional (3-D) space. combination two 3-D datasets provided fine-scale reconstruction characteristics adjacent within forests. A reconstruction, assembled from nine Echidna Validation...