• Journal of Korea Multimedia Society
• /
• v.13 no.5
• /
• pp.792-801
• /
• 2010

In this paper, we introduce a novel scene graph library for mobile platforms, called the 'Mobile OpenSceneGraph (Mobile OSG)', as a mobile graphics middleware. Our mobile scene graph library supports several nice properties, including platform-independence, extensibility, touch-based UI supports, and compatibility, by carefully adapting the OpenSceneGraph library, one of the most widely used graphics middlewares for desktop platforms, to mobile platforms. We validate the usefulness of the proposed library for mobiles with the several experimental results including real-time rendering, camera manipulations with a touch-based UI, and animations with switching geometric nodes.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.5
• /
• pp.205-212
• /
• 2019

Scene graph is a kind of knowledge graph that represents both objects and their relationships found in a image. This paper proposes a 3D scene graph generation model for three-dimensional indoor environments. An 3D scene graph includes not only object types, their positions and attributes, but also three-dimensional spatial relationships between them, An 3D scene graph can be viewed as a prior knowledge base describing the given environment within that the agent will be deployed later. Therefore, 3D scene graphs can be used in many useful applications, such as visual question answering (VQA) and service robots. This proposed 3D scene graph generation model consists of four sub-networks: object detection network (ObjNet), attribute prediction network (AttNet), transfer network (TransNet), relationship prediction network (RelNet). Conducting several experiments with 3D simulated indoor environments provided by AI2-THOR, we confirmed that the proposed model shows high performance.

• Proceedings of the Korea Multimedia Society Conference
• /
• 1998.10a
• /
• pp.155-160
• /
• 1998

VRML 파서는 VRML(.wrl) 파일을 읽기 위하여 반드시 필요한 부분으로서, 이미 공개되어 있는 것으로는 C/C++ 코드의 QvLib1.0과 Yacc & Lex로 개발한 VRML2.0 파서가 있다. 그러나 QvLib1.0의 경우 VRML2.0 파일에 적용할 수 없으며, Yacc & Lex로 개발한 파서는 생성된 소스 코드를 수정하기 어려울 뿐만 아니라 이를 수정하기 위해서는 자동 생성기인 Yacc & Lex의 사용 방법을 잘 알고 있어야 한다. 이에 본 논문에서는 C/C++ 코드의 VRML2.0 파서를 설계 및 구현하고 파싱의 결과로 생성된 parse tree를 이용하여 VRML 파일을 scene graph로 bygus할 수 있도록 하는 Scene Graph 생성기의 구현과정을 소개하였다. 또한 본 연구의 파서와 Scene Graph 생성기를 기 개발된 VRML 저작도구에 적용 및 테스트하였다.

• Journal of Information Processing Systems
• /
• v.16 no.6
• /
• pp.1250-1260
• /
• 2020

This study proposes a novel deep neural network model that can accurately detect objects and their relationships in an image and represent them as a scene graph. The proposed model utilizes several multimodal features, including linguistic features and visual context features, to accurately detect objects and relationships. In addition, in the proposed model, context features are embedded using graph neural networks to depict the dependencies between two related objects in the context feature vector. This study demonstrates the effectiveness of the proposed model through comparative experiments using the Visual Genome benchmark dataset.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.9 no.1
• /
• pp.11-16
• /
• 2014

When it comes to develop flight simulator, HAT (Height Above Terrain) is required to provide altitude information to the pilot who learns how to control an airplane in landing and takeoff situation. However, there might be inconsistent problem between real terrain and simulation information since current implementation of HAT simply depends on center of gravity point on the airplane. To overcome mentioned problem, in this paper, we propose how to obtain more accurate altitude information than existing scheme by making use of HAT and HOT (Height Of Terrain) information of landing equipments according to movement of the airplane. Moreover, we demonstrate the accuracy of the proposed scheme through new flight simulator developed through OSG(OpenSceneGraph) by taking example of terrain information for domestic airport.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.12
• /
• pp.3383-3397
• /
• 2023

Scene graphs serve as semantic abstractions of images and play a crucial role in enhancing visual comprehension and reasoning. However, the performance of Scene Graph Generation is often compromised when working with biased data in real-world situations. While many existing systems focus on a single stage of learning for both feature extraction and classification, some employ Class-Balancing strategies, such as Re-weighting, Data Resampling, and Transfer Learning from head to tail. In this paper, we propose a novel approach that decouples the feature extraction and classification phases of the scene graph generation process. For feature extraction, we leverage a transformer-based architecture and design an adaptive calibration function specifically for predicate classification. This function enables us to dynamically adjust the classification scores for each predicate category. Additionally, we introduce a Distribution Alignment technique that effectively balances the class distribution after the feature extraction phase reaches a stable state, thereby facilitating the retraining of the classification head. Importantly, our Distribution Alignment strategy is model-independent and does not require additional supervision, making it applicable to a wide range of SGG models. Using the scene graph diagnostic toolkit on Visual Genome and several popular models, we achieved significant improvements over the previous state-of-the-art methods with our model. Compared to the TDE model, our model improved mR@100 by 70.5% for PredCls, by 84.0% for SGCls, and by 97.6% for SGDet tasks.

• Journal of KIISE:Software and Applications
• /
• v.31 no.12
• /
• pp.1611-1620
• /
• 2004

MPEG-4 supports dynamic scene composition through add/delete/replace of object or change of object's properties. Other existing MPEG-4 players focus on transmitting and playing the multimedia data according to MPEG-4 standard. It is insufficient for MPEG-4's characteristic such as playback of various objects and playback of dynamic scene composition. In this paper, we propose an effective scene compositor which is the core component of MPEG-4 player The scene compositor is an optimized processor that searches efficiently the scene graph, creates the data structure for independent management of object information and improves processing ability of user interaction. The scene compositor supports sufficiently scene description information, and is managed independently in player for component extension and application of mobile environment.

• The Journal of Korea Robotics Society
• /
• v.9 no.1
• /
• pp.1-10
• /
• 2014

In recent years, the research of 3D mapping technique in urban environments obtained by mobile robots equipped with multiple sensors for recognizing the robot's surroundings is being studied actively. However, the map generated by simple integration of multiple sensors data only gives spatial information to robots. To get a semantic knowledge to help an autonomous mobile robot from the map, the robot has to convert low-level map representations to higher-level ones containing semantic knowledge of a scene. Given a 3D point cloud of an urban scene, this research proposes a method to recognize the objects effectively using 3D graph model for autonomous mobile robots. The proposed method is decomposed into three steps: sequential range data acquisition, normal vector estimation and incremental graph-based segmentation. This method guarantees the both real-time performance and accuracy of recognizing the objects in real urban environments. Also, it can provide plentiful data for classifying the objects. To evaluate a performance of proposed method, computation time and recognition rate of objects are analyzed. Experimental results show that the proposed method has efficiently in understanding the semantic knowledge of an urban environment.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.04a
• /
• pp.457-459
• /
• 2012

가상현실 환경은 사용자가 데이터와 상호작용하면서 직관적으로 데이터를 인지할 수 있게 해주는 인터페이스를 제공한다. 이런 가상현실 환경은 고해상도 디스플레이 장치에서 사실상 유일하게 사용할 수 있는 사용자 친화적인 인터페이스 방식이기도 하다. 본 논문에서는 고해상도 디스플레이 장치를 이용한 몰입형 가상현실 환경에서 사용되는 사용자 인터페이스 프레임워크를 소개하기로 한다. 이 인터페이스 프레임워크는 VR Juggler와 OSG(OpenSceneGraph)를 기반으로 하며, 다양한 형태의 디스플레이 장치를 이용한 몰입형 가상현실 환경에서 사용자가 가상현실과 상호작용하는데 필요한 기반 환경을 제공한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2011.11a
• /
• pp.455-458
• /
• 2011

컴퓨터에서 생성된 시뮬레이션 결과는 가시화(Visualization)라는 과정을 통해 인간이 직관적으로 해석하기 쉬운 형태로 가공된다. 사람이 직관적으로 이해하기 어려운 수치의 나열로 구성돼 있던 데이터가 컴퓨터 그래픽스 기술의 적용을 통해 보다 쉽게 분석할 수 있는 형태로 변형되는 것이다. 최근에는 이런 시뮬레이션 데이터의 크기와 복잡도가 점점 증가하고 있는 추세로, 이런 추세에 따라 고해상도 디스플레이 장치나 몰입형 가상현실 장치의 필요성은 점점 증가하고 있다. 이런 이유로 클러스터 시스템을 이용한 고해상도의 디스플레이 장치에서 해당 시스템에 대한 전문적인 지식을 갖추지 않은 일반 연구자가 상호작용할 수 있는 사용자 인터페이스에 대한 필요성도 높아지고 있다. 본 논문에서는 고해상도 디스플레이 장치에서 데이터를 가시화하는 프레임워크와 그 프레임워크에서 사용하는 통합 환경 인터페이스를 소개하기로 한다. 이 인터페이스는 VR Juggler와 OSG(OpenSceneGraph)를 기반으로 하며, 다양한 형태의 디스플레이 장치와 몰입형 가상현실 환경에서 시뮬레이션 데이터와 실시간 상호작용을 통한 해석을 수행하는데 필요한 기반 환경을 제공한다.