• International journal of advanced smart convergence
• /
• v.7 no.2
• /
• pp.95-100
• /
• 2018

In this paper, we propose a binarized multi-scale module to accelerate the speed of the pose estimating deep neural network. Recently, deep learning is also used for fine-tuned tasks such as pose estimation. One of the best performing pose estimation methods is based on the usage of two neural networks where one computes the heat maps of the body parts and the other computes the part affinity fields between the body parts. However, the convolution filtering with a large kernel filter takes much time in this model. To accelerate the speed in this model, we propose to change the large kernel filters with binarized multi-scale modules. The large receptive field is captured by the multi-scale structure which also prevents the dropdown of the accuracy in the binarized module. The computation cost and number of parameters becomes small which results in increased speed performance.

• The Journal of Korea Robotics Society
• /
• v.18 no.2
• /
• pp.182-188
• /
• 2023

Due to the acceleration of an aging society, the need for lower limb exoskeletons to assist gait is increasing. And for use in daily life, it is essential to have technology that can accurately estimate gait phase even in the walking environment and walking speed of the wearer that changes frequently. In this paper, we implement an LSTM-based gait phase estimation learning model by collecting gait data according to changes in gait speed in outdoor level ground and stair environments. In addition, the results of the gait phase estimation error for each walking environment were compared after learning for both max hip extension (MHE) and max hip flexion (MHF), which are ground truth criteria in gait phase divided in previous studies. As a result, the average error rate of all walking environments using MHF reference data and MHE reference data was 2.97% and 4.36%, respectively, and the result of using MHF reference data was 1.39% lower than the result of using MHE reference data.

• The Journal of Korea Robotics Society
• /
• v.14 no.1
• /
• pp.1-7
• /
• 2019

Recently, smart factories have attracted much attention as a result of the 4th Industrial Revolution. Existing factory automation technologies are generally designed for simple repetition without using vision sensors. Even small object assemblies are still dependent on manual work. To satisfy the needs for replacing the existing system with new technology such as bin picking and visual servoing, precision and real-time application should be core. Therefore in our work we focused on the core elements by using deep learning algorithm to detect and classify the target object for real-time and analyzing the object features. We chose YOLO CNN which is capable of real-time working and combining the two tasks as mentioned above though there are lots of good deep learning algorithms such as Mask R-CNN and Fast R-CNN. Then through the line and inside features extracted from target object, we can obtain final outline and estimate object posture.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.7
• /
• pp.17-24
• /
• 2009

Increasing use of information system has led to larger amount of developing expenses and demands on software. Until recent days, the model using regression analysis based on statistical algorithm has been used. However, Machine learning is more investigated now. This paper estimates the software cost using SVR(Support Vector Regression). a sort of machine learning technique. Also, it finds the best set of parameters applying immune algorithm. In this paper, software cost estimation is performed by SVR based on immune algorithm while changing populations, memory cells, and number of allele. Finally, this paper analyzes and compares the result with existing other machine learning methods.

• Journal of KIISE:Software and Applications
• /
• v.36 no.11
• /
• pp.960-966
• /
• 2009

A manifold is used to represent a relationship between high-dimensional data samples in low-dimensional space. In human pose estimation, it is created in low-dimensional space for processing image and 3D body configuration data. Manifold learning is to build a manifold. But it is vulnerable to silhouette variations. Such silhouette variations are occurred due to view-change, person-change, distance-change, and noises. Representing silhouette variations in a single manifold is impossible. In this paper, we focus a silhouette variation problem occurred by view-change. In previous view invariant pose estimation methods based on manifold learning, there were two ways. One is modeling manifolds for all view points. The other is to extract view factors from mapping functions. But these methods do not support one by one mapping for silhouettes and corresponding body configurations because of unsupervised learning. Modeling manifold and extracting view factors are very complex. So we propose a method based on triple manifolds. These are view manifold, pose manifold, and body configuration manifold. In order to build manifolds, we employ biased manifold learning. After building manifolds, we learn mapping functions among spaces (2D image space, pose manifold space, view manifold space, body configuration manifold space, 3D body configuration space). In our experiments, we could estimate various body poses from 24 view points.

• Journal of the Korea Management Engineers Society
• /
• v.23 no.4
• /
• pp.159-164
• /
• 2018

Human-machine serial systems must be normal in both systems. Though the failure of machine is irreducible by itself, the human errors are of recurring type. When the human performance is described quantitatively, non-homogeneous Poisson Process model of human errors can be developed. And the model parameters can be estimated by maximum likelihood estimation and numerical analysis method. System reliability is obtained by multiplying machine reliability by human reliability.

• Journal of the Korean Society for Precision Engineering
• /
• v.16 no.8
• /
• pp.186-192
• /
• 1999

This paper introduces a study of a method for the forward kinematic analysis, which finds the 6 DOF motions and velocities from the given six cylinder lengths in the Stewart platform. From the viewpoints of kinematics, the solution for the inverse kinematic is easily found by using the vectors of the links which are composed of the joint coordinates in base and plate frames, to act contrary to the serial manipulator, but forward kinematic is difficult because of the nonlinearity and complexity of the Stewart platform dynamic equation with the multi-solutions. Hence we, first in this study, introduce the linear estimator using the Luenberger's observer, and the estimator using the nonlinear measured model for the forward kinematic solutions. But it is difficult to find the parameter of the design for the estimation gain or to select the estimation gain and the constant steady state error exists. So this study suggests the estimator with the estimation gain to be learned by the neural network with the structure of multi-perceptron and the learning method using back propagation and shows the estimation performance using the simulation.

• Journal of Platform Technology
• /
• v.11 no.5
• /
• pp.16-22
• /
• 2023

There is a growing interest in facial age estimation because many applications require age estimation techniques from facial images. In order to estimate the exact age of a face, a technique for extracting aging features from a face image and classifying the age according to the extracted features is required. Recently, the performance of various CNN-based deep learning models has been greatly improved in the image recognition field, and various CNN-based deep learning models are being used to improve performance in the field of facial age estimation. In this paper, age estimation performance was compared by learning facial features based on various CNN-based models such as AlexNet, VGG-16, VGG-19, ResNet-18, ResNet-34, ResNet-50, ResNet-101, ResNet-152. As a result of experiment, it was confirmed that the performance of the facial age estimation models using ResNet-34 was the best.

• Journal of IKEEE
• /
• v.28 no.1
• /
• pp.6-13
• /
• 2024

This paper proposes an approach to detect abnormal data in automotive controller area network (CAN) using an unsupervised learning model, i.e. autoencoder and Gaussian kernel density estimation function. The proposed autoencoder model is trained with only message ID of CAN data frames. Afterwards, by employing the Gaussian kernel density estimation function, it effectively detects abnormal data based on the trained model characterized by the optimally determined number of frames and a loss threshold. It was verified and evaluated using four types of attack data, i.e. DoS attacks, gear spoofing attacks, RPM spoofing attacks, and fuzzy attacks. Compared with conventional unsupervised learning-based models, it has achieved over 99% detection performance across all evaluation metrics.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.8 no.4
• /
• pp.25-34
• /
• 2012

In this study, in the process of testing before the release of the software products designed, software testing manager in advance should be aware of the testing-information. Therefore, the effective learning effects perspective has been studied using the NHPP software. The finite failure nonhomogeneous Poisson process models presented and applied property of learning effect based on truncated time and delayed S-shaped software reliability. Software error detection techniques known in advance, but influencing factors for considering the errors found automatically and learning factors, by prior experience, to find precisely the error factor setting up the testing manager are presented comparing the problem. As a result, the learning factor is greater than autonomous errors-detected factor that is generally efficient model can be confirmed. This paper, a failure data analysis was performed, using time between failures, according to the small sample and large sample sizes. The parameter estimation was carried out using maximum likelihood estimation method. Model selection was performed using the mean square error and coefficient of determination, after the data efficiency from the data through trend analysis was performed.