• The Transactions of the Korea Information Processing Society
• /
• v.13 no.7
• /
• pp.335-347
• /
• 2024

Open-vocabulary 3D point cloud instance segmentation (OV-3DIS) is a challenging visual task to segment a 3D scene point cloud into object instances of both base and novel classes. In this paper, we propose a novel model Open3DME for OV-3DIS to address important design issues and overcome limitations of the existing approaches. First, in order to improve the quality of class-agnostic 3D masks, our model makes use of T3DIS, an advanced Transformer-based 3D point cloud instance segmentation model, as mask proposal module. Second, in order to obtain semantically text-aligned visual features of each point cloud segment, our model extracts both 2D and 3D features from the point cloud and the corresponding multi-view RGB images by using pretrained CLIP and OpenSeg encoders respectively. Last, to effectively make use of both 2D and 3D visual features of each point cloud segment during label assignment, our model adopts a unique feature ensemble method. To validate our model, we conducted both quantitative and qualitative experiments on ScanNet-V2 benchmark dataset, demonstrating significant performance gains.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.3
• /
• pp.8-15
• /
• 2007

Recently, it is important to process multimedia data efficiently. Especially, in case of retrieval of multimedia information, technique of user interface and retrieval technique are necessary. This paper proposes a new technique which detects cuts effectively in compressed image information by MPEG. A cut is a turning point of scenes. The cut-detection is the basic work and the first-step for video indexing and retrieval. Existing methods have a weak point that they detect wrong cuts according to change of a screen such as fast motion of an object, movement of a camera and a flash. Because they compare between previous frame and present frame. The proposed technique detects shots at first using DC(Direct Current) coefficient of DCT(Discrete Cosine Transform). The database is composed of these detected shots. Features are extracted by HMMD color model and edge histogram descriptor(EHD) among the MPEG-7 visual descriptors. And detections are performed in sequence by the proposed matching technique. Through this experiments, an improved video segmentation system is implemented that it performs more quickly and precisely than existing techniques have.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.6
• /
• pp.315-325
• /
• 2016

Due to the expansion of data and technical progress in the military industry, it is important to extract meaningful information for assuring quality and making policies. The analysis of trends and decision making based on big data is helpful for increasing productivity in business and finding new business opportunities. We propose an application to collect reliable quality information for munitions and build a big data platform for using the accumulated information and numerical data. We verified the proposed platform using the Test Report Information Service (TRIS) system and suggest a method that utilizes unstructured and semi-structured data accumulated by TRIS. Thus, we expect that the proposed platform will help in building infrastructure for military data, making efficient strategies, and analyzing trends for assuring munitions quality.

• 한국HCI학회:학술대회논문집
• /
• 2007.02b
• /
• pp.136-144
• /
• 2007

1990년대 후반까지 줄곧 광고 시장 점유율 1위를 차지하던 신문광고가 2001년을 기점으로 광고 시장 점유율의 지속적인 하락세를 보이고 있다. 이는 인터넷과 같은 경쟁적인 뉴미디어가 급격히 성장하는 데 그 원인이 있을 뿐만 아니라, 신문의 광고효과를 산출하는 과학적 방법론이 부족하다는 데에도 그 원인이 있다. 이러한 문제의식에서 출발하여, 본 연구에서는 안구운동 추적(eyetracking)을 활용하여 신문광고의 노출효과를 측정했다. 이를 위하여, 기존 신문 구독자의 인구통계학적 분포에 따라 추출된 50명의 표본을 대상으로, 신문의 면내 광고별 체류시간을 헤드 마운트형 안구운동 추적기(head-mounted eyetracker)를 활용하여 측정했다. 실험 결과, 데이터의 95%가 기존의 주의-기억 단계를 따르는 순차적인 정보처리모델과 스타치 구독률 조사(starch readership reports) 결과와 안구운동 추적 결과와 일치했다. 하지만, 5%의 데이터가 주의와 기억 단계의 차이에 기인한 이월 효과(carry-out)로 순차적인 정보처리모델과 일치하지 않는 것으로 나타났다. 즉 5%가량이 광고를 보지 않았음에도 봤다고 표시했다. 또한, 기존에 스타치 구독률 조사와 같은 재인 측정 방법을 통하여 간접적으로 측정하였던 기존의 신문 광고 노출율(보통 30%)에 비하여, 직접 측정한 결과 60%에 달하는 광고 노출율을 보였다. 이는 기존의 인식과 달리, 스타치 구독률 조사가 단순 노출 정도가 아니라, 응답자의 선택적인 기억 결과를 측정한 것임을 보여준다. 이로써, 안구운동 추적을 활용한 노출효과측정은 기존에 스타치 구독률 조사에서 놓치고 있는 부분을 보완할 수 있는 기법임을 보여준다. 또한 이 기법에 의하여 도출된 노출율은 60%에 달하여 기존의 약30%로 제시된 연구결과들이 실제로 측정기법의 한계로 인하여, 기억이 가지고 있는 왜곡을 그대로 반영한 노출율에 관한 자료를 쓰는 것으로 나타났다. 본 연구에서 제안된 노출효과 측정방법론은 기존의 왜곡된 데이터를 보완하고, 신문 광고 시장의 과학적인 측정기법을 발전시키는 데 의미 있는 기여를 할 것으로 기대한다.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.6
• /
• pp.561-568
• /
• 2011

The purpose of this study is to measure displacements that occurs on a surface and interior of slope model and the shape when the slope is destroyed at vibration experiment of the slope model using close range photogrammetry. The circle targets and sphere targets are installed on a chamber and a slope model, while the earthquake wave are applied in regular time interval. The close range photogrammetric images are acquired in each displacements step until the slope model is destroyed. Those photos are processed by image processing method and the center points of targets are automatically extracted. Furthermore, the three-dimensional coordinates of targets are calculated by image orientation and bundle adjustment processing. As a result, amount of displacement at each level is precisely measured and provided the basic information for assessing the slope stability using three-dimensional measurement of the target movement and slope destruction.

• Journal of Engineering Education Research
• /
• v.12 no.2
• /
• pp.96-106
• /
• 2009

Problem-based learning has been considered as one of the effective educational methods in engineering education. However, in so far as professors who require practical insights in PBL and experiences of developing actual problems by subject, in particular, thorough understanding from experiences of PBL process as well as problem-development has not been sufficiently provided. The purpose of this paper is to present strategies focusing on problem design for PBL on engineering education. In order to do this, a literature review and a qualitative case study were performed. Especially, the study intended to identify differences and gap between professors' problems-development process and its output and those of authentic PBL. Professors were found that their PBL problems had lack of authenticity, consideration on experiences of students, and realistic thinking process. Professors in PBL had difficulty to link theory into real situation. In consequence, in designing a problem, we consider the followings; first, the problem should be designed based on real design process and its output. Second, the problem should be designed and implemented in all academic years for developing student's systematic and skillful thinking process. In conclusion, more supports are needed for engineering professors to extend their experiences of designing and developing actual problems that present real experience.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.17 no.6
• /
• pp.89-98
• /
• 2007

We have come a long way in the information age. Thanks to the advancement of such technologies as the internet, we have discovered new ways to convey information on a broader scope. However, negative aspects exist as is with anything else. These may include invasion of privacy over the web, or identity theft over the internet. What is more alarming is that malwares so called 'maliciouscodes' are rapidly spreading. Its intent is very destructive which can result in hacking, phishing and as aforementioned, one of the most disturbing problems on the net, invasion of privacy. This thesis describes the technology of how you can effectively analyze and detect these kind of malicious codes. We propose sequencial hybrid analysis for API calls that are hooked inside user-mode and kernel-level of Windows. This research explains how we can cope with malicious code more efficiently by abstracting malicious function signature and hiding attribute.

• Journal of Digital Convergence
• /
• v.11 no.2
• /
• pp.243-248
• /
• 2013

This paper proposes an effective method using Adaptive Boosting to track a person's head in complex background. By only one way to feature extraction methods are not sufficient for modeling a person's head. Therefore, the method proposed in this paper, several feature extraction methods for the accuracy of the detection head running at the same time. Feature Extraction for the imaging of the head was extracted using sub-region and Haar wavelet transform. Sub-region represents the local characteristics of the head, Haar wavelet transform can indicate the frequency characteristics of face. Therefore, if we use them to extract the features of face, effective modeling is possible. In the proposed method to track down the man's head from the input video in real time, we ues the results after learning Harr-wavelet characteristics of the three types using AdaBoosting algorithm. Originally the AdaBoosting algorithm, there is a very long learning time, if learning data was changes, and then it is need to be performed learning again. In order to overcome this shortcoming, in this research propose efficient method using cascade AdaBoosting. This method reduces the learning time for the imaging of the head, and can respond effectively to changes in the learning data. The proposed method generated classifier with excellent performance using less learning time and learning data. In addition, this method accurately detect and track head of person from a variety of head data in real-time video images.

• The KIPS Transactions:PartB
• /
• v.10B no.4
• /
• pp.443-450
• /
• 2003

In this paper, we proposed the novel hierarchical algorithm for the recognition of English calling cards that processes multiresolution images of calling cards hierarchically to extract individual characters and recognizes the extracted characters by using the enhanced neural network method. The hierarchical recognition algorithm generates multiresolution images of calling cards, and each processing step in the algorithm selects and processes the image with suitable resolution for lower processing overhead and improved output. That is, first, the image of 1/3 times resolution, to which the horizontal smearing method is applied, is used to extract the areas including only characters from the calling card image, and next, by applying the vertical smearing and the contour tracking masking, the image of a half time resolution is used to extract individual characters from the character string areas. Lastly, the original image is used in the recognition step, because the image includes the morphological information of characters accurately. And for the recognition of characters with diverse font types and various sizes, the enhanced RBF network that improves the middle layer based on the ART1 was proposed and applied. The results of experiments on a large number of calling card images showed that the proposed algorithm is greatly improved in the performance of character extraction and recognition compared with the traditional recognition algorithms.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.47 no.6
• /
• pp.86-96
• /
• 2010

In this paper, we propose stereo vision-based obstacle detection and vehicle verification methods using U-disparity map and bird's-eye view mapping. First, we extract a road feature using maximum frequent values in each row and column. And we extract obstacle areas on the road using the extracted road feature. To extract obstacle areas exactly we utilize U-disparity map. We can extract obstacle areas exactly on the U-disparity map using threshold value which consists of disparity value and camera parameter. But there are still multiple obstacles in the extracted obstacle areas. Thus, we perform another processing, namely segmentation. We convert the extracted obstacle areas into a bird's-eye view using camera modeling and parameters. We can segment obstacle areas on the bird's-eye view robustly because obstacles are represented on it according to ranges. Finally, we verify the obstacles whether those are vehicles or not using various vehicle features, namely road contacting, constant horizontal length, aspect ratio and texture information. We conduct experiments to prove the performance of our proposed algorithms in real traffic situations.