• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.2
• /
• pp.73-81
• /
• 2024

Recently, mobile manipulators have been utilized in agriculture industry for weed removal and harvest automation. This paper proposes a domain adaptive fruit detection method for harvest automation, by utilizing OWL-ViT model which is an open-vocabulary object detection model. The vision-language model can detect objects based on text prompt, and therefore, it can be extended to detect objects of undefined categories. In the development of deep learning models for real-world problems, constructing a large-scale labeled dataset is a time-consuming task and heavily relies on human effort. To reduce the labor-intensive workload, we utilized a large-scale public dataset as a source domain data and employed a domain adaptation method. Adversarial learning was conducted between a domain discriminator and feature extractor to reduce the gap between the distribution of feature vectors from the source domain and our target domain data. We collected a target domain dataset in a real-like environment and conducted experiments to demonstrate the effectiveness of the proposed method. In experiments, the domain adaptation method improved the AP50 metric from 38.88% to 78.59% for detecting objects within the range of 2m, and we achieved 81.7% of manipulation success rate.

• International Journal of Advanced Culture Technology
• /
• v.12 no.3
• /
• pp.148-158
• /
• 2024

We investigate Autism Spectrum Disorder (ASD), which is typified by deficits in social interaction, repetitive behaviors, limited vocabulary, and cognitive delays. Traditional diagnostic methodologies, reliant on expert evaluations, frequently result in deferred detection and intervention, particularly in South Korea, where there is a dearth of qualified professionals and limited public awareness. In this study, we employ advanced deep learning algorithms to enhance early ASD screening through automated video analysis. Utilizing architectures such as Convolutional Long Short-Term Memory (ConvLSTM), Long-term Recurrent Convolutional Network (LRCN), and Convolutional Neural Networks with Gated Recurrent Units (CNN+GRU), we analyze video data from platforms like YouTube and TikTok to identify stereotypic behaviors (arm flapping, head banging, spinning). Our results indicate that the LRCN model exhibited superior performance with 79.61% accuracy on the augmented platform video dataset and 79.37% on the original SSBD dataset. The ConvLSTM and CNN+GRU models also achieved higher accuracy than the original SSBD dataset. Through this research, we underscore AI's potential in early ASD detection by automating the identification of stereotypic behaviors, thereby enabling timely intervention. We also emphasize the significance of utilizing expanded datasets from social media platform videos in augmenting model accuracy and robustness, thus paving the way for more accessible diagnostic methods.

• Journal of the Korean Society for Library and Information Science
• /
• v.56 no.4
• /
• pp.5-28
• /
• 2022

In this study, LD construction cases of overseas libraries were analyzed with focus on published datasets, reused vocabulary, and interlinked external datasets, and based on the analysis results, basic data on LD construction plans of domestic libraries were obtained. As a result of the analysis of 21 library cases, overseas libraries have established a faithful authority LD and conducted new services using published LDs. To this end, overseas libraries collaborated with other libraries and cultural institutions within the region, within the country, and nationally under the leadership of the library, and based on this cooperation, a specialized dataset was published. Overseas libraries used Schema.org to increase the visibility of published LDs, and used BIBFRAME for subdivision of description to define various entities and build LDs based on the defined entities. Overseas libraries have utilized various defined entities to link related information, display results, browse, and download in bulk. Overseas libraries were interested in the continuous up-to-date of interlinked external datasets, and directly utilized external data to reinforce catalog information. In this study, based on the derived implications, points to be considered when issuing LDs by domestic libraries were proposed. The research results can be used as basic data when future domestic libraries plan LD services or upgrade existing services.

• Journal of Information Processing Systems
• /
• v.19 no.3
• /
• pp.355-369
• /
• 2023

Since the widespread adoption of deep-learning and related distributed representation, there have been substantial advancements in part-of-speech (POS) tagging for many languages. When training word representations, morphology and shape are typically ignored, as these representations rely primarily on collecting syntactic and semantic aspects of words. However, for tasks like POS tagging, notably in morphologically rich and resource-limited language environments, the intra-word information is essential. In this study, we introduce a deep neural network (DNN) for POS tagging that learns character-level word representations and combines them with general word representations. Using the proposed approach and omitting hand-crafted features, we achieve 90.47%, 80.16%, and 79.32% accuracy on our own dataset for three morphologically rich languages: Uyghur, Uzbek, and Kyrgyz. The experimental results reveal that the presented character-based strategy greatly improves POS tagging performance for several morphologically rich languages (MRL) where character information is significant. Furthermore, when compared to the previously reported state-of-the-art POS tagging results for Turkish on the METU Turkish Treebank dataset, the proposed approach improved on the prior work slightly. As a result, the experimental results indicate that character-based representations outperform word-level representations for MRL performance. Our technique is also robust towards the-out-of-vocabulary issues and performs better on manually edited text.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.411-414
• /
• 2021

One of the elements of technology that has become extremely critical within the field of education today is Deep learning. It has been especially used in the area of natural language processing, with some word-representation vectors playing a critical role. However, some of the low-resource languages, such as Swahili, which is spoken in East and Central Africa, do not fall into this category. Natural Language Processing is a field of artificial intelligence where systems and computational algorithms are built that can automatically understand, analyze, manipulate, and potentially generate human language. After coming to discover that some African languages fail to have a proper representation within language processing, even going so far as to describe them as lower resource languages because of inadequate data for NLP, we decided to study the Swahili language. As it stands currently, language modeling using neural networks requires adequate data to guarantee quality word representation, which is important for natural language processing (NLP) tasks. Most African languages have no data for such processing. The main aim of this project is to recognize and focus on the classification of words in English, Swahili, and Korean with a particular emphasis on the low-resource Swahili language. Finally, we are going to create our own dataset and reprocess the data using Python Script, formulate the syllabic alphabet, and finally develop an English, Swahili, and Korean word analogy dataset.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.415-417
• /
• 2021

Since the words and words in Chinese sentences are continuous and the length of vocabulary is huge, Chinese NER(Named Entity Recognition) always based on character representation. In recent years, many Chinese research has been reconsidered how to integrate the word information into the Chinese NER model. However, the traditional sequence model has complex structure, the slow inference speed, and an additional dictionary information is needed, which is difficult to implement in the industry. The approach in this paper has the state of the art and parallelizable, which is integrated the char-word embeddings, so that the model learns word information. The proposed model is easy to implement, and outperforms traditional model in terms of speed and efficiency, which is improved f1-score on two dataset.

• Journal of Information Science Theory and Practice
• /
• v.12 no.3
• /
• pp.63-74
• /
• 2024

This study aimed to develop a digital thesaurus dedicated to cataloging the traditional common culture of the Greater Mekong Subregion. The process followed a meticulous seven-step methodology, including scoping, vocabulary collection, knowledge structure analysis, relationship delineation, related word adjustments, list validation, and evaluation. Leveraging principles from knowledge organization, thesaurus construction, and digital platform development, the TemaTres web application emerged as the primary tool for constructing this thesaurus. The study's results showed that 2,042 principal words related to the traditional common culture of the Greater Mekong Subregion were compiled and classified into terms for each of the seven deep levels. Each term was accompanied by essential metadata, including broader and narrower terms, related terms, cross-references, and scope notes. This rich dataset empowered semantic search capabilities across diverse applications and web services, providing access to knowledge pertaining to the traditional common culture of the Greater Mekong Subregion and contributing to a deeper understanding of this cultural domain.

• 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 Korean Society for information Management
• /
• v.33 no.3
• /
• pp.321-343
• /
• 2016

This study aimed to design a semi-automatic web-based pilot system 1) to build a Korean novel geo-name, 2) to update the database using automatic geo-name extraction for a scalable database, and 3) to retrieve/visualize the usage of an old geo-name on the map. In particular, the problem of extracting novel geo-names, which are currently obsolete, is difficult to solve because obtaining a corpus used for training dataset is burden. To build a corpus for training data, an admin tool, HTML crawler and parser in Python, crawled geo-names and usages from a vocabulary dictionary for Korean New Novel enough to train a named entity tagger for extracting even novel geo-names not shown up in a training corpus. By means of a training corpus and an automatic extraction tool, the geo-name database was made scalable. In addition, the system can visualize the geo-name on the map. The work of study also designed, implemented the prototype and empirically verified the validity of the pilot system. Lastly, items to be improved have also been addressed.

• Journal of KIISE:Software and Applications
• /
• v.36 no.12
• /
• pp.1034-1037
• /
• 2009

Sign language spotting is the task of detecting the start and end points of signs from continuous data and recognizing the detected signs in the predefined vocabulary. The difficulty with sign language spotting is that instances of signs vary in both motion and shape. Moreover, signs have variable motion in terms of both trajectory and length. Especially, variable sign lengths result in problems with spotting signs in a video sequence, because short signs involve less information and fewer changes than long signs. In this paper, we propose a method for spotting variable lengths signs based on semi-CRF (semi-Markov Conditional Random Field). We performed experiments with ASL (American Sign Language) and KSL (Korean Sign Language) dataset of continuous sign sentences to demonstrate the efficiency of the proposed method. Experimental results show that the proposed method outperforms both HMM and CRF.