• Journal of KIBIM
• /
• v.14 no.2
• /
• pp.13-24
• /
• 2024

This research suggests a novel visualization approach utilizing Generative AI to render photorealistic architectural alternatives images in the early design phase. Photorealistic rendering intuitively describes alternatives and facilitates clear communication between stakeholders. Nevertheless, the conventional rendering process, utilizing 3D modelling and rendering engines, demands sophisticate model and processing time. In this context, the paper suggests a rendering approach employing the text-to-image method aimed at generating a broader range of intuitive and relevant reference images. Additionally, it employs an Text-to-Image method focused on producing a diverse array of alternatives reflecting architects' styles when visualizing the exteriors of residential buildings from the mass model images. To achieve this, fine-tuning for architects' styles was conducted using the Low-Rank Adaptation (LoRA) method. This approach, supported by fine-tuned models, allows not only single style-applied alternatives, but also the fusion of two or more styles to generate new alternatives. Using the proposed approach, we generated more than 15,000 meaningful images, with each image taking only about 5 seconds to produce. This demonstrates that the Generative AI-based visualization approach significantly reduces the labour and time required in conventional visualization processes, holding significant potential for transforming abstract ideas into tangible images, even in the early stages of design.

• International conference on construction engineering and project management
• /
• 2024.07a
• /
• pp.1285-1285
• /
• 2024

In process plant construction, the implementation of design automation technologies is pivotal in reducing the timeframes associated with the design phase and in enabling the generation and evaluation of a variety of design alternatives, thereby facilitating the identification of optimal solutions. These technologies can play a crucial role in ensuring the successful delivery of projects. Previous research in the domain of design automation has primarily focused on parametric design in architectural contexts and on the automation of equipment layout and pipe routing within plant engineering, predominantly employing rule-based algorithms. Nevertheless, these studies are constrained by the limited flexibility of their models, which narrows the scope for generating alternative solutions and complicates the process of exploring comprehensive solutions using nonlinear optimization techniques as the number of design and engineering parameters increases. This research introduces a framework for automating plant design through the use of generative neural network models to overcome these challenges. The framework is applicable to the layout problems of process plants, covering the equipment necessary for production processes and the facilities for essential resources and their interconnections. The development of the proposed Neural-network (NN) based Generative Design Model unfolds in four stages: (a) Rule-based Model Development: This initial phase involves the development of rule-based models for layout generation and evaluation, where the generation model produces layouts based on predefined parameters, and the evaluation model assesses these layouts using various performance metrics. (b) Neural Network Model Development: This phase transitions towards neural network models, establishing a NN-based layout generation model utilizing Generative Adversarial Network (GAN)-based methods and a NN-based layout evaluation model. (c) Model Optimization: The third phase is dedicated to optimizing the models through Bayesian Optimization, aiming to extend the exploration space beyond the limitations of rule-based models. (d) Inverse Design Model Development: The concluding phase employs an inverse design method to merge the generative and evaluative networks, resulting in a model that outputs layout designs to meet specific performance objectives. This study aims to augment the efficiency and effectiveness of the design process in process plant construction, transcending the limitations of conventional rule-based approaches and contributing to the achievement of successful project outcomes.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.28 no.1
• /
• pp.1-21
• /
• 2024

This review comprehensively explores the evolution, theoretical underpinnings, variations, and applications of diffusion models. Originating as a generative framework, diffusion models have rapidly ascended to the forefront of machine learning research, owing to their exceptional capability, stability, and versatility. We dissect the core principles driving diffusion processes, elucidating their mathematical foundations and the mechanisms by which they iteratively refine noise into structured data. We highlight pivotal advancements and the integration of auxiliary techniques that have significantly enhanced their efficiency and stability. Variants such as bridges that broaden the applicability of diffusion models to wider domains are introduced. We put special emphasis on the ability of diffusion models as a crucial foundation model, with modalities ranging from image, 3D assets, and video. The role of diffusion models as a general foundation model leads to its versatility in many of the downstream tasks such as solving inverse problems and image editing. Through this review, we aim to provide a thorough and accessible compendium for both newcomers and seasoned researchers in the field.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.5
• /
• pp.1590-1609
• /
• 2021

Laughter is one of the most important nonverbal sound that human generates. It is a means for expressing his emotions. The acoustic and contextual features of this specific sound are different from those of speech and many difficulties arise during their modeling process. During this work, we propose an audio laughter generation system based on unsupervised generative models: the autoencoder (AE) and its variants. This procedure is the association of three main sub-process, (1) the analysis which consist of extracting the log magnitude spectrogram from the laughter database, (2) the generative models training, (3) the synthesis stage which incorporate the involvement of an intermediate mechanism: the vocoder. To improve the synthesis quality, we suggest two hybrid models (LSTM-VAE, GRU-VAE and CNN-VAE) that combine the representation learning capacity of variational autoencoder (VAE) with the temporal modelling ability of a long short-term memory RNN (LSTM) and the CNN ability to learn invariant features. To figure out the performance of our proposed audio laughter generation process, objective evaluation (RMSE) and a perceptual audio quality test (listening test) were conducted. According to these evaluation metrics, we can show that the GRU-VAE outperforms the other VAE models.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.8
• /
• pp.223-230
• /
• 2024

The purposes of this study is to find out how to use Generative AI by class stage and class model so that classes can be planned using various Generative AI in elementary and secondary education. To this end, contents of using Generative AI according to general instructional stages and instructional models by school level and subject were developed, and revised and supplemented through review by 13 field experts. As for the method of using Generative AI by class stage, general class stages were divided into three stages: 'class preparation', 'in class', and 'class arrangement', and the subject of using Generative AI at each stage, the contents of using it, and the types of Generative AI that can be used are summarized. As a method of using Generative AI according to the class model, eight class contents were developed based on teaching and learning models according to the characteristics of each school level and subject. In order to expand the use of Generative AI in elementary and secondary classes, it is necessary to develop more diverse class contents by school level and subject and distribute them in the field. It is also necessary to develop educational materials on matters to consider when using Generative AI in class.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.3
• /
• pp.233-238
• /
• 2014

Creativity is crucial in designing and producing attractive accessaries and daily supplies as well as art works. Generative design can be a paradigm to be used to obtain novel ideas or motifs for creative design works. This paper introduces a generative design method which comes up with unique ring models using genetic algorithm. It presents how the genetic algorithm works in terms of candidate solution coding, operators, and fitness evaluation function. The proposed method allows the customers to express their personal preference and later the preference to be reflected in fitness evaluation. In the final stage of the proposed method, several ring models are suggested for customers to choose on their own. The chosen ring models can be put into physical rings with the help of a 3D printer because the models are expressed in 3D geometric structures.

• Journal of The Korean Association For Science Education
• /
• v.16 no.1
• /
• pp.13-34
• /
• 1996

The purpose of this study is to develop appropriate science teaching models which can be applied effectively to relevant situations. Five science teaching models; cognitive conflict teaching models, generative teaching model, learning cycle teaching model, hypothesis verification teaching model and discovery teaching model, were identified from the existing models. The teaching models were modified and in primary and secondary students using a nonequivalent pretest-posttest control group design. Major findings of this study were as follows: 1. For teaching science concepts, three teaching models were found more effective; cognitive conflict teaching model, generative teaching model and discovery teaching model. 2. For teaching inquiry skills, two teaching models were found more effective; learning cycle teaching model and hypothesis verification teaching model. 3. For teaching scientific attitudes, two teaching models were found more effective; learning cycle teaching models and discovery teaching model. Each teaching model requires specific learning environment. It is strongly suggested that teachers should select a suitable teaching model carefully after evaluating the learning environment including teacher and student variables, learning objectives and curricular materials.

• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.37-42
• /
• 2019

In this paper, we explore the details of three classic data augmentation methods and two generative model based oversampling methods. The three classic data augmentation methods are random sampling (RANDOM), Synthetic Minority Over-sampling Technique (SMOTE), and Adaptive Synthetic Sampling (ADASYN). The two generative model based oversampling methods are Conditional Generative Adversarial Network (CGAN) and Wasserstein Generative Adversarial Network (WGAN). In imbalanced data, the whole instances are divided into majority class and minority class, where majority class occupies most of the instances in the training set and minority class only includes a few instances. Generative models have their own advantages when they are used to generate more plausible samples referring to the distribution of the minority class. We also adopt CGAN to compare the data augmentation performance with other methods. The experimental results show that WGAN-based oversampling technique is more stable than other approaches (RANDOM, SMOTE, ADASYN and CGAN) even with the very limited training datasets. However, when the imbalanced ratio is too small, generative model based approaches cannot achieve satisfying performance than the conventional data augmentation techniques. These results suggest us one of future research directions.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.13 no.3
• /
• pp.186-199
• /
• 2013

We pose pattern classification as a density estimation problem where we consider mixtures of generative models under partially labeled data setups. Unlike traditional approaches that estimate density everywhere in data space, we focus on the density along the decision boundary that can yield more discriminative models with superior classification performance. We extend our earlier work on the recursive estimation method for discriminative mixture models to semi-supervised learning setups where some of the data points lack class labels. Our model exploits the mixture structure in the functional gradient framework: it searches for the base mixture component model in a greedy fashion, maximizing the conditional class likelihoods for the labeled data and at the same time minimizing the uncertainty of class label prediction for unlabeled data points. The objective can be effectively imposed as individual mixture component learning on weighted data, hence our mixture learning typically becomes highly efficient for popular base generative models like Gaussians or hidden Markov models. Moreover, apart from the expectation-maximization algorithm, the proposed recursive estimation has several advantages including the lack of need for a pre-determined mixture order and robustness to the choice of initial parameters. We demonstrate the benefits of the proposed approach on a comprehensive set of evaluations consisting of diverse time-series classification problems in semi-supervised scenarios.

• Journal of the Korea Convergence Society
• /
• v.13 no.1
• /
• pp.13-21
• /
• 2022

Recent deep learning-based research shows excellent performance in most natural language processing (NLP) fields with pre-trained language models. In particular, the auto-encoder-based language model proves its excellent performance and usefulness in various fields of Korean language understanding. However, the decoder-based Korean generative model even suffers from generating simple sentences. Also, there is few detailed research and data for the field of conversation where generative models are most commonly utilized. Therefore, this paper constructs multi-turn dialogue data for a Korean generative model. In addition, we compare and analyze the performance by improving the dialogue ability of the generative model through transfer learning. In addition, we propose a method of supplementing the insufficient dialogue generation ability of the model by extracting recommended response candidates from external knowledge information through a retrival model.