• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.168-173
• /
• 2007

게임이나 가상환경에서 오락성과 실감성을 증진시키는 여러 가지 방법들 가운데 지능적인 Non-Player Character (NPC)들의 존재는 중요하다. 컴퓨터 그래픽과 관련 하드웨어 플랫폼 기술의 발전으로 인해 사용자들은 이제 시각적인 만족을 넘어서서, NPC들이 보다 지능적으로 행동하면서 오락적인 만족감과 동시에 보다 향상된 실감성을 제공하기를 원한다. 하지만, 유한상태기계 (Finite State Machine, FSM)를 기반으로 하는 NPC 구현의 한계와 어려움으로 인해 이러한 사용자들의 요구사항을 만족시키는 것은 어렵다. 본 논문에서는 FSM과 추론 엔진(Inference Engine)을 결합한 새로운 NPC 행동제어 구조를 제안한다. 또한 제안된 구조의 가능성을 시연하기 위해 실제로 동작하는 데모를 소개한다. 이러한 FSM과 추론 엔진의 하이브리드 구조는 FSM이 제공하는 NPC 반응의 실시간성을 보장하는 동시에 추론 엔진이 제공할 수 있는 보다 지능적이고 계획적인 NPC들의 행동을 만들어 낼 수 있다는 장점이 있다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2017.01a
• /
• pp.155-156
• /
• 2017

본 논문에서는 블리자드 회사에서 제작한 게임 중 하나인 오버워치(Overwatch)에서 게임의 간접정보를 학습하여 플레이어에게 유리한 영웅 캐릭터를 추천해주는 인공신경망 기반 영웅 캐릭터 추천 방법을 제안한다. 오버워치에서 게임 맵별로 적군 캐릭터와 아군 캐릭터가 선정되었을 때 플레이어가 어떤 영웅캐릭터를 선정하면 승률에 좋은지를 알기가 어렵다. 본 논문에서는 플레이어의 영웅캐릭터 선정을 도와주기위하여 오버워치 게임의 간접정보를 기반으로 학습데이터를 만들어 인공신경망을 학습한 후 학습한 인공신경망을 이용하여 영웅캐릭터를 추천한다. 실험결과 인공신경망이 추천하는 영웅캐릭터가 적절한 캐릭터임을 확인하였다.

• Journal of Forest and Environmental Science
• /
• v.32 no.2
• /
• pp.182-188
• /
• 2016

In this study 'forest design in waterside unused land' of 'waterside area' in the specificity that multilateral, including landscape, recreation, social and environmental aspects in the space, taking into account by multiple factors :Raise 'practicality' than a functional management and police to hang out with, each element is usually way to improve the beauty. There is currently carried out waterside green area creation, which is artificial plantation and open spaces only emphasizing 'landscape' with lacking multi-function purpose, as a 'forest design in waterside unused land' strategy but it should be lowered by performance in this strategy. And the study suggested a strategy which prepared cultural and social infrastructure to be able to attract different fund and build local character as an alternative demand plan considered environmental character as a top priority.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09b
• /
• pp.41-44
• /
• 2003

인공생명 기법의 대표적인 응용분야 중 하나인 창발적인 개체의 생성에서 유전자를 표현하는 방법으로 크게 방향성 그래프와 L-system의 두 가지 방법이 사용되고 있다. 두 가지 방법 모두 각기 좋은 특성이 있지만 L-system은 사용자에 의해 내부적인 파라메터로 정의되기 때문에 전체적인 실제 모양을 구조적으로 정의하기 어렵다 본 논문에서는 이런 단점을 극복하고자 실제 모양을 표현하는데 적절한 구조적 방향성 그래프를 도입하여 개체의 유전자를 구조적으로 표현함으로써 실제 개체와 유사한 모양을 창발적으로 생성하고, 이를 클러스터 대화형 유전자 알고리즘에 적용하여 사용자가 생성하고자 하는 실제 모양의 캐릭터를 자동으로 생성하게 하였다. 실험을 통해 2차원에서의 의자모양과 꽃 모양을 생성함으로써 다양한 구조의 창발적인 캐릭터를 생성할 수 있는 가능성을 제시했다.

• Journal of the Semiconductor & Display Technology
• /
• v.13 no.1
• /
• pp.57-62
• /
• 2014

This paper proposes a new optical recognition method of credit card numbers. Firstly, the proposed method segments numbers from the input image of a credit card. It uses the significant differences of standard deviations between the foreground numbers and the background. Secondly, the method extracts gradient features from the segmented numbers. The gradient features are defined as four directions of grayscale pixels for 16 regions of an input number. Finally, it utilizes an artificial neural network classifier that uses an error back-propagation algorithm. The proposed method is implemented using C language in an embedded Linux system for a high-speed real-time image processing. Experiments were conducted by using real credit card images. The results show that the proposed algorithm is quite successful for most credit cards. However, the method fails in some credit cards with strong background patterns.

• Journal of information and communication convergence engineering
• /
• v.21 no.2
• /
• pp.167-173
• /
• 2023

Hangul is unique compared to other Asian languages because of its simple letter forms that combine to create syllabic shapes. There are 24 basic letters that can be combined to form 27 additional complex letters. This produces 51 graphemes. Hangul optical character recognition has been a research topic for some time; however, handwritten Hangul recognition continues to be challenging owing to the various writing styles, slants, and cursive-like nature of the handwriting. In this study, a dataset containing thousands of samples of 51 Hangul graphemes was gathered from 110 freshmen university students to create a robust dataset with high variance for training an artificial neural network. The collected dataset included 2200 samples for each consonant grapheme and 1100 samples for each vowel grapheme. The dataset was normalized to the MNIST digits dataset, trained in three neural networks, and the obtained results were compared.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.11
• /
• pp.1782-1793
• /
• 1989

Stroke crossing number, syntactic pattern recognition procedure, top down recognition structure, and heuristic approach are studied for the Korean text recognition. We propose new algorithms: 1)Korean vowel seperation using limited scanning method in the Korean characters, 2) extracting strokes using stroke width method, 3) stroke crossing number and its properties, 4) average, standard deviation, and mode of stroke crossing number, and 5) classification and recognition methods of limited chinese character. These are studied with computer simuladtions and experiments.

• Proceedings of the KIEE Conference
• /
• 1999.07e
• /
• pp.2420-2422
• /
• 1999

To live comportably, proper air and lighting environment are necessary, environment is obtained by excuting. We can get the lighting environment with excuting day-lighting and artificial lighting suittably. The developing trend of fluorescent light lamp call for low power and high lighting efficiency. Nowadays, The 26 mm FL is developed and in wide use in this proper, the character of 26 mm FL is discussed, same improvement are proposed.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.3
• /
• pp.111-119
• /
• 2010

An artificial light source has a character that a light emits from a point to all directions with many radial and straight ray shapes. It is very difficult and complex to render those emitting lights. Also, users have difficulty in expressing exactly 3D objects because of colors varying with changing of a light and having many parameters. In this paper, to solve those problems we design an artificial light source using X3D to create a model that represents easily many radial and straight ray shapes and propose the online system that each factors of colors to be reflected by a light is separated and then users can control them to detect object's colors by a mouse. Various light sources with reality can be easily created using proposed system.

• ETRI Journal
• /
• v.44 no.3
• /
• pp.476-490
• /
• 2022

With the recent advances in technology, automatic speech recognition (ASR) has been widely used in real-world applications. The efficiency of converting large amounts of speech into text accurately with limited resources has become more vital than ever. In this study, we propose a method to rapidly recognize a large speech database via a transformer-based end-to-end model. Transformers have improved the state-of-the-art performance in many fields. However, they are not easy to use for long sequences. In this study, various techniques to accelerate the recognition of real-world speeches are proposed and tested, including decoding via multiple-utterance-batched beam search, detecting end of speech based on a connectionist temporal classification (CTC), restricting the CTC-prefix score, and splitting long speeches into short segments. Experiments are conducted with the Librispeech dataset and the real-world Korean ASR tasks to verify the proposed methods. From the experiments, the proposed system can convert 8 h of speeches spoken at real-world meetings into text in less than 3 min with a 10.73% character error rate, which is 27.1% relatively lower than that of conventional systems.