• Journal of Information Processing Systems
• /
• v.11 no.2
• /
• pp.248-265
• /
• 2015

In this paper, we propose a maximum entropy-based model, which can mathematically explain the bio-molecular event extraction problem. The proposed model generates an event table, which can represent the relationship between an event trigger and its arguments. The complex sentences with distinctive event structures can be also represented by the event table. Previous approaches intuitively designed a pipeline system, which sequentially performs trigger detection and arguments recognition, and thus, did not clearly explain the relationship between identified triggers and arguments. On the other hand, the proposed model generates an event table that can represent triggers, their arguments, and their relationships. The desired events can be easily extracted from the event table. Experimental results show that the proposed model can cover 91.36% of events in the training dataset and that it can achieve a 50.44% recall in the test dataset by using the event table.

• Water Engineering Research
• /
• v.5 no.1
• /
• pp.47-54
• /
• 2004

Several computer programs have been developed to make stochastically generated weather data from observed daily data. But they require fully dataset to run WGEN. Mostly, meterological data frequently have sporadic missing data as well as totally missing data. The modified WGEN has data filling algorithm for incomplete meterological datasets. Any other WGEN models have not the function of data filling. Modified WGEN with data filling algorithm is processing from the equation of Matalas for first order autoregressive process on a multi dimensional state with known cross and auto correlations among state variables. The parameters of the equation of Matalas are derived from existing dataset and derived parameters are adopted to fill data. In case of WGEN (Richardson and Wright, 1984), it is one of most widely used weather generators. But it has to be modified and added. It uses an exponential distribution to generate precipitation amounts. An exponential distribution is easier to describe the distribution of precipitation amounts. But precipitation data with using exponential distribution has not been expressed well. In this paper, generated precipitation data from WGEN and Modified WGEN were compared with corresponding measured data as statistic parameters. The modified WGEN adopted a formula of CLIGEN for WEPP (Water Erosion Prediction Project) in USDA in 1985. In this paper, the result of other parameters except precipitation is not introduced. It will be introduced through study of verification and review soon

• Journal of Environmental Science International
• /
• v.19 no.10
• /
• pp.1229-1235
• /
• 2010

Three meteor-statistical forecasting models - the transfer function model, the time-series autoregressive model and the neural networks model - were tested to develop a daily forecasting model for Jejudo, where the need and demand for wind power forecasting has increased. All the meteorological observation sites in Jejudo have been classified into 6 groups using a cluster analysis. Four pairs of observation sites among them, all having strong wind speed correlation within the same meteorological group, were chosen for a model test. In the development of the wind speed forecasting model for Jejudo, it was confirmed that not only the use a wind dataset at the objective site itself, but the introduction of another wind dataset at the nearest site having a strong wind speed correlation within the same group, would enhance the goodness to fit of the forecasting. A transfer function model and a neural network model were also confirmed to offer reliable predictions, with the similar goodness to fit level.

• ETRI Journal
• /
• v.43 no.4
• /
• pp.603-616
• /
• 2021

Over the last few years, autonomous vehicles have progressed very rapidly. The odometry technique that estimates displacement from consecutive sensor inputs is an essential technique for autonomous driving. In this article, we propose a fast, robust, and accurate odometry technique. The proposed technique is light detection and ranging (LiDAR)-based direct odometry, which uses a spherical range image (SRI) that projects a three-dimensional point cloud onto a two-dimensional spherical image plane. Direct odometry is developed in a vision-based method, and a fast execution speed can be expected. However, applying LiDAR data is difficult because of the sparsity. To solve this problem, we propose an SRI generation method and mathematical analysis, two key point sampling methods using SRI to increase precision and robustness, and a fast optimization method. The proposed technique was tested with the KITTI dataset and real environments. Evaluation results yielded a translation error of 0.69%, a rotation error of 0.0031°/m in the KITTI training dataset, and an execution time of 17 ms. The results demonstrated high precision comparable with state-of-the-art and remarkably higher speed than conventional techniques.

• ETRI Journal
• /
• v.44 no.6
• /
• pp.1004-1019
• /
• 2022

For autonomous driving without high-definition maps, we present a model capable of generating multiple plausible paths from egocentric images for autonomous vehicles. Our generative model comprises two neural networks: feature extraction network (FEN) and path generation network (PGN). The FEN extracts meaningful features from an egocentric image, whereas the PGN generates multiple paths from the features, given a driving intention and speed. To ensure that the paths generated are plausible and consistent with the intention, we introduce an attentive discriminator and train it with the PGN under a generative adversarial network framework. Furthermore, we devise an interaction model between the positions in the paths and the intentions hidden in the positions and design a novel PGN architecture that reflects the interaction model for improving the accuracy and diversity of the generated paths. Finally, we introduce ETRIDriving, a dataset for autonomous driving, in which the recorded sensor data are labeled with discrete high-level driving actions, and demonstrate the state-of-the-art performance of the proposed model on ETRIDriving in terms of accuracy and diversity.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.521-523
• /
• 2021

Data augmentation are techniques used to increase the amount of data by using small amount of existing data. With the spread of the Internet, we can easily obtain data. However, there are still certain industries, like medicine, where it is difficult to obtain data. The same is true for image data in which a black screen is displayed on video wall controller. Because it is rare that a black screen is displayed during operation, it is not easy to obtain an image with a black screen. We propose a DCGAN based architecture that generate dataset using a small amount of black screen image.

• Journal of Korean Society for Geospatial Information Science
• /
• v.13 no.1 s.31
• /
• pp.55-61
• /
• 2005

Digital Elevation Model (DEM) generation from remotely sensed imagery is crucial for a variety of mapping applications such as ortho-photo generation, city modeling. High resolution imaging satellites such as SPOT-5, IKONOS, QUICK-BIRD, ORBVIEW constitute an excellent source for efficient and economic generation of DEM data. However, prerequisite knowledge in the areas of sensor modeling, epipolar resampling, and image matching is required to generate DEM from these high resolution satellite imagery. From the above requirements, epipolar resampling emerges as the most important factors. Research attempts in this area are still in high demand and short supply. Another cause that adds to the complication of the problem is that most studies of DEM generation from IKONOS scenes have been based on rational function model. In this paper, we proposed a new methodology for DEM generation from satellite scenes using parallel projection model which is sensor independent, makes it possible for sensor modeling and epipolar resampling by only few control points. The performance and feasibility of the developed methodology is evaluated through real dataset captured by IKONOS.

• Journal of the Korean GEO-environmental Society
• /
• v.19 no.1
• /
• pp.5-14
• /
• 2018

All kinds of monitoring data in construction site could have outlier created from diverse cause. In this study generation technique of synthesis value, its regression, final outlier detection and assessment are conducted to distinct outlier data included in extensive time series dataset. Synthesis value having weight factor of correlation between a number of datasets consist of many monitoring data enable to detect outlier by increasing its correlation. Standard artificial dataset in which intentional outliers are inserted has been used for assessment of synthesis value technique. These results showed increase of detection accuracy for outlier and general tendency in case of having different time series models in common. Accuracy of outlier detection increased in case of using more dataset and showing similar time series pattern.

• The KIPS Transactions:PartB
• /
• v.13B no.2 s.105
• /
• pp.133-138
• /
• 2006

Learning with Attribute Value Taxonomies (AVT) has shown that it is possible to construct accurate, compact and robust classifiers from a partially missing dataset (dataset that contains attribute values specified with different level of precision). Yet, in many cases AVTs are generated from experts or people with specialized knowledge in their domain. Unfortunately these user-provided AVTs can be time-consuming to construct and misguided during the AVT building process. Moreover experts are occasionally unavailable to provide an AVT for a particular domain. Against these backgrounds, this paper introduces an AVT generating method called GA-AVT-Learner, which finds a near optimal AVT with a given training dataset using a genetic algorithm. This paper conducted experiments generating AVTs through GA-AVT-Learner with a variety of real world datasets. We compared these AVTs with other types of AVTs such as HAC-AVTs and user-provided AVTs. Through the experiments we have proved that GA-AVT-Learner provides AVTs that yield more accurate and compact classifiers and improve performance in learning missing data.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.5
• /
• pp.969-976
• /
• 2019

A lightweight artificial intelligence hardware has made great strides in many application areas. In general, a lightweight artificial intelligence system consist of lightweight artificial intelligence engine and preprocessor including feature selection, generation, extraction, and normalization. In order to achieve optimal performance in broad range of applications, lightweight artificial intelligence system needs to choose a good preprocessing function and set their respective hyper-parameters. This paper proposes a unified framework for a lightweight artificial intelligence system and utilization method for finding models with optimal performance to use on a given dataset. The proposed unified framework can easily generate a model combined with preprocessing functions and lightweight artificial intelligence engine. In performance evaluation using handwritten image dataset and fall detection dataset measured with inertial sensor, the proposed unified framework showed building optimal artificial intelligence models with over 90% test accuracy.