• KIISE Transactions on Computing Practices
• /
• v.24 no.3
• /
• pp.138-144
• /
• 2018

A video is a vivid medium similar to human's visual-linguistic experiences, since it can inculcate a sequence of situations, actions or dialogues that can be told as a story. In this study, we propose story learning/regeneration frameworks from videos with successive event order supervision for contextual coherence. The supervision induces each episode to have a form of trajectory in the latent space, which constructs a composite representation of ordering and semantics. In this study, we incorporated the use of kids videos as a training data. Some of the advantages associated with the kids videos include omnibus style, simple/explicit storyline in short, chronological narrative order, and relatively limited number of characters and spatial environments. We build the encoder-decoder structure with successive event order embedding, and train bi-directional LSTMs as sequence models considering multi-step sequence prediction. Using a series of approximately 200 episodes of kids videos named 'Pororo the Little Penguin', we give empirical results for story regeneration tasks and SEOE. In addition, each episode shows a trajectory-like shape on the latent space of the model, which gives the geometric information for the sequence models.

• Journal of Korean Society of Transportation
• /
• v.20 no.5
• /
• pp.67-79
• /
• 2002

In this paper, observed trajectories of a vehicle platoon are viewed as one realization of a finite sequence of random trajectories. In this point of view, we develop novel and mathematically rigorous concept of traffic flow variables such as local traffic density, instantaneous traffic flow, and velocity field and investigate their nature on a general probability space of a sequence of random trajectories which represent vehicle trajectories. We present a simple model of random trajectories as an illustrative example and, derive the values of traffic flow variables based on the new definitions in this model. In particular, we construct the model for the sequence of random vehicle trajectories with a system of stochastic differential equations. Each equation of the system nay represent microscopic random maneuvering behavior of each vehicle with properly designed drift coefficient functions and diffusion coefficient functions. The system of stochastic differential equations nay generate a well-defined probability space of a sequence of random vehicle trajectories. We derive the partial differential equation for the expected cumulative plot with appropriate initial conditions. By solving the equation with numerical methods, we obtain the values of expected cumulative plot, local traffic density, and instantaneous traffic flow. In addition, we derive the partial differential equation for the expected travel time to a certain location with appropriate initial and/or boundary conditions, which is solvable numerically. We apply this model to a case of single vehicle trajectory.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.30 no.7
• /
• pp.51-60
• /
• 2002

The Monte-Carlo simulation of statistical analysis is used to investigate the final conditions of states as well as the footprint boundaries resulting from the atmospheric re-entry dispersions. The re-entry dispersions in this paper are specified by a $7\times7$ covariance matrix of latitude, longitude, altitude, bank angle, flight path angle, heading error, and range at entry velocity. The error sources that affect these at re-entry for a deboost are the uncertainties associated with atmospheric density and temperature, initial errors, wind, and estimation error of aerodynamic coefficients. Using $3{\sigma}_n$ deviations of these errors and a nominal flight trajectory, the covariance matrix of state variables can be determined by performing a trajectory error analysis. Major considerations in the application of the Monte-Carlo method are the simulation of perturbed trajectories, bank reversal, and determination of the impact points for each of these trajectories. This paper analyzes the results of uncertainties from the viewpoint of aero-coefficients and bank reversal.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.42 no.2
• /
• pp.158-164
• /
• 2014

LEO(Low Earth Orbit) Satellite which is discarded should be reentered to atmosphere in 25 years by '25 years rule' of IADC(Inter-Agency Space Debris Coordination Committee) Guidelines. If the parts of satellite are survived from severe aerothermodynamic condition, it could damage to human and property. South Korea operates KOMPSAT-2 and STSAT series as LEO satellite. It is necessary to dispose of them by reentering atmosphere. Therefore this paper analyze the trajectory, survivability, casualty area and casualty probability of a virtual LEO satellite using ESA(European Space Agency)'s DRAMA(Debris Risk Assesment and Mitigation Analysis) tool. As a result, it is noted that casuality area is $15.2742m^2$ and casualty probability is 5.9614E-03 then will be survived 198.831kg.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.41 no.4
• /
• pp.314-318
• /
• 2013

Tests and evaluations following the U.S. MIL-HDBK/STANDARD were successfully conducted to assure the performance of the air-to-surface guidance kit which was developed first in Korea. Various ground tests confirmed the operation capability and reliability of the guidance kit, and flight tests proved very good mid-range gliding performance and accuracy of the gliding bomb which was a general purpose bomb with the guidance kit.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.1
• /
• pp.6-11
• /
• 2014

Repetitive control is a specialized control scheme to track and/or attenuate a periodic reference trajectory and/or disturbance. Most researches about repetitive control have been performed in the frequency domain. Recently, several approaches to deal with repetitive control systems in the state space are developed by representing a q filter as a state-space equation. This paper presents a design method of a repetitive control system in the state space to satisfy $H_{\infty}$ performance. The overall system is composed of a plant, a repetitive controller, and a state-feedback controller, which can be converted to a standard form used in $H_{\infty}$ control. A LMI (Linear Matrix Inequality)-based stability condition is derived for fixed state-feedback gains. Under a given q filter, another LMI condition is derived to improve $H_{\infty}$ performance and is employed to find state-feedback gains by solving an optimization problem. Finally, to verify the feasibility of the proposed method, a numerical example is demonstrated.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.659-663
• /
• 2009

In this paper, we propose a rectification method that can convert ray space data obtained by controlled camera array to ideal data. Here, Ideal data is obtained by getting longitudinal and transversal epipolar line between cameras vertical and horizontal. However it is actually difficult to arrange cameras strictly because we arrange cameras by hand. As conventional method, we have use camera-calibration method. But if we use this method there are some errors on the output image. When we generate arbitrary viewpoint images this error is critical problem. We focus attention on ideal trajectory of characteristic point. And to minimize the error directly we parallelize the real one. And we showed usefulness of proposed technique. Then using the proposed technique, we were successful reducing the error to less than 0.5 pixels.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.42 no.3
• /
• pp.213-220
• /
• 2014

A program that predicts trajectories of projectiles influenced by the interference flow field of helicopters is developed. The interference flow field are computed using a compressible inviscid solver in conjunction with an actuator disc model. The trajectories are predicted using 6-DOF (Degree of Freedom) equations as well as an alternative form of modified point mass equations of motion. The method for the interference flow field prediction method are validated with ROBIN(ROtor Body INteraction) model. A Sierra international bullet and a 105mm projectile are used to validate the trajectory method. Trajectories of a Sierra International bullet and a HYDRA 70 rocket firing from a helicopter are predicted.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.46 no.5
• /
• pp.385-392
• /
• 2018

This paper suggests a new method to estimate the roll and pitch damping moment coefficients of a flying disc through sensor data from the onboard IMU module. This method can be easily performed than wind tunnel or computational fluid dynamics methods because it estimates aerodynamic coefficients simply after accumulating the inertial data through several repeated flight experiments. Estimated coefficients are applied to a simulator which is based on the flight dynamics of a flying disc. Finally, the predicted flight trajectory is compared with the true position provided by GPS, which demonstrated the validity of the proposed estimation method.

• International Journal of Aeronautical and Space Sciences
• /
• v.14 no.4
• /
• pp.369-378
• /
• 2013

This paper presents a vision/LiDAR integrated navigation system that provides accurate relative navigation performance on a general ground surface, in GNSS-denied environments. The considered ground surface during flight is approximated as a piecewise continuous model, with flat and slope surface profiles. In its implementation, the presented system consists of a strapdown IMU, and an aided sensor block, consisting of a vision sensor and a LiDAR on a stabilized gimbal platform. Thus, two-dimensional optical flow vectors from the vision sensor, and range information from LiDAR to ground are used to overcome the performance limit of the tactical grade inertial navigation solution without GNSS signal. In filter realization, the INS error model is employed, with measurement vectors containing two-dimensional velocity errors, and one differenced altitude in the navigation frame. In computing the altitude difference, the ground slope angle is estimated in a novel way, through two bisectional LiDAR signals, with a practical assumption representing a general ground profile. Finally, the overall integrated system is implemented, based on the extended Kalman filter framework, and the performance is demonstrated through a simulation study, with an aircraft flight trajectory scenario.