• Journal of the Korea Computer Graphics Society
• /
• v.26 no.3
• /
• pp.87-97
• /
• 2020

In this paper, we propose a novel neural relighting that infers a relighted rendering image based on the user-guided specular highlight map. The proposed network utilizes a pre-trained neural renderer as a backbone network learned from the rendered image of a 3D scene with various lighting conditions. We jointly optimize a 3D light position and its associated relighted image by back-propagation, so that the difference between the base image and the relighted image is similar to the user-guided specular highlight map. The proposed method has the advantage of being able to explicitly infer the 3D lighting position, while providing the artists' preferred 2D screen-space interface. The performance of the proposed network was measured under the conditions that can establish ground truths, and the average error rate of light position estimations is 0.11, with the normalized 3D scene size.

• Journal of the Korea Computer Graphics Society
• /
• v.27 no.4
• /
• pp.11-24
• /
• 2021

In this paper, we propose an artificial neural network framework that can represent the foam effects expressed in liquid simulation in detail without noise. The position and advection of foam particles are calculated using the existing screen projection method, and the noise problem that appears in this process is solved through an proposed artificial neural network. The important thing in the screen projection approach is the projection map, but noise occurs in the projection map in the process of projecting momentum into the discretized screen space, and we efficiently solve this problem by using an artificial neural network-based denoising network. When the foam generating area is selected through the projection map, 2D is inversely transformed into 3D space to generate foam particles. We solve the existing denoising network problem in which small-scaled foam particles disappear. In addition, by integrating the proposed algorithm with the screen-space projection framework, all the advantages of this approach can be accommodated. As a result, it shows through various experiments whether it is possible to stably represent not only the clean foam effects but also the foam particles lost due to the denoising process.

• Journal of Korea Multimedia Society
• /
• v.22 no.6
• /
• pp.655-664
• /
• 2019

In this paper, we propose a novel way of combining multiple deep convolutional neural network (DCNN) architectures which work well for accurate video face identification by adopting a serial combination of 3D and 2D DCNNs. The proposed method first divides an input video sequence (to be recognized) into a number of sub-video sequences. The resulting sub-video sequences are used as input to the 3D DCNN so as to obtain the class-confidence scores for a given input video sequence by considering both temporal and spatial face feature characteristics of input video sequence. The class-confidence scores obtained from corresponding sub-video sequences is combined by forming our proposed class-confidence matrix. The resulting class-confidence matrix is then used as an input for learning 2D DCNN learning which is serially linked to 3D DCNN. Finally, fine-tuned, serially combined DCNN framework is applied for recognizing the identity present in a given test video sequence. To verify the effectiveness of our proposed method, extensive and comparative experiments have been conducted to evaluate our method on COX face databases with their standard face identification protocols. Experimental results showed that our method can achieve better or comparable identification rate compared to other state-of-the-art video FR methods.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.41.2-41
• /
• 2002

The stereo correspondence of two retinal images is one of the most difficult problems in stereo vision because the reconstruction of 3-D scene is a typical visual ill-posed problem. So far there still have been many unsolved problems, one of which is to reconstruct 3-D scene for a monochromatic surface because there is no clue to make a correspondence between two retinal images. We consider this problem with two layered self-organization neural network to simulate the competitive and cooperative interaction of binocular neurons. A...

• Journal of the Korean Society for Precision Engineering
• /
• v.20 no.4
• /
• pp.112-117
• /
• 2003

This paper deals with the problem of synthesis of the 3-dimensional Grasp Planning. In previous studies the genetic algorithm has been used to find optimal grasping points, but it had a limitation such as the determination time of grasping points was so long. To overcome this limitation we proposed a new algorithm which employs the Neural Network. In the Neural network we chose input parameters based on the shape of the object and output parameters resulted from optimization with the GA method. In this study the GRNN method is employed, it has been trained by the result value of optimization method and it has been tested by known object. The algorithm is verified by computer simulation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.2
• /
• pp.311-326
• /
• 2024

The rapid development of neural network technology promotes the neural network model driven by big data to overcome the texture effect of complex objects. Due to the limitations in complex scenes, it is necessary to establish custom template matching and apply it to the research of many fields of computational vision technology. The dependence on high-quality small label sample database data is not very strong, and the machine learning system of deep feature connection to complete the task of texture effect inference and speculation is relatively poor. The style transfer algorithm based on neural network collects and preserves the data of patterns, extracts and modernizes their features. Through the algorithm model, it is easier to present the texture color of patterns and display them digitally. In this paper, according to the texture effect reasoning of custom template matching, the 3D visualization of the target is transformed into a 3D model. The high similarity between the scene to be inferred and the user-defined template is calculated by the user-defined template of the multi-dimensional external feature label. The convolutional neural network is adopted to optimize the external area of the object to improve the sampling quality and computational performance of the sample pyramid structure. The results indicate that the proposed algorithm can accurately capture the significant target, achieve more ablation noise, and improve the visualization results. The proposed deep convolutional neural network optimization algorithm has good rapidity, data accuracy and robustness. The proposed algorithm can adapt to the calculation of more task scenes, display the redundant vision-related information of image conversion, enhance the powerful computing power, and further improve the computational efficiency and accuracy of convolutional networks, which has a high research significance for the study of image information conversion.

• Journal of Korea Multimedia Society
• /
• v.12 no.3
• /
• pp.451-459
• /
• 2009

This paper presents an approach to detect premature ventricular contraction(PVC) using discrete wavelet transform and fuzzy neural network. As the input of the algorithm, we use 14 coefficients of d3, d4, and d5, which are transformed by a discrete wavelet transform(DWT). This paper uses a neural network with weighted fuzzy membership functions(NEWFM) to diagnose PVC. The NEWFM discussed in this paper classifies a normal beat and a PVC beat. The size of the window of DWT is $-31/360{\sim}+32/360$ second(64 samples) whose center is the R wave. Using the seven records of the MIT-BIH arrhythmia database used in Shyu's paper, the classification performance of the proposed algorithm is 99.91%, which outperforms the 97.04% of Shyu's analysis. Using the forty records of the M1T-BIH arrhythmia database used in Inan's paper, the classification performance of the proposed algorithm is 98.01%, which outperforms 96.85% of Inan's one. The SE and SP of the proposed algorithm are 84.67% and 99.39%, which outperforms the 82.57% and 98.33%, respectively, of Inan's study.

• 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.

• Transactions of Materials Processing
• /
• v.8 no.4
• /
• pp.331-339
• /
• 1999

In this study, to analyze the shaped drawing process from round bar, the practical conical die with considering die radius and bearing was defined by a mathematical expression, and also a simple technique for initial mesh generation to the shaped drawing process was proposed. The drawing of rectangular section from round bar, one of the shaped drawing process, has been simulated by using non-steady state 3D rigid plastic finite element method in order to evaluate the influence of semi-die angle and reduction in area to corner filling. Other process variables such as friction constant, rectangular ratio, die radius and bearing length were fixed during the simulation. An artificial neural network has been introduced to obtain the optimal process conditions which gave rise to a fast simulation.

• The Journal of the Korea Contents Association
• /
• v.9 no.7
• /
• pp.19-26
• /
• 2009

This paper presents an approach to classify normal and ventricular tachycardia/fibrillation(VT/VF) from the Creighton University Ventricular Tachyarrhythmia Database(CUDB) using the neural network with weighted fuzzy membership functions(NEWFM) and wavelet transforms. In the first step, wavelet transforms are used to obtain the detail coefficients at levels 3 and 4. In the second step, all of detail coefficients d3 and d4 are classified into four intervals, respectively, and then the standard deviations of the specific intervals are used as eight numbers of input features of NEWFM. NEWFM classifies normal and VT/VF beats using eight numbers of input features, and then the accuracy rate is 90.1%.