• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.8
• /
• pp.1111-1120
• /
• 2022

Image Capturing is a matter of viewing images and describing images in language. The problem is an important problem that can be solved by keeping, understanding, and bringing together two areas of image processing and natural language processing. In addition, by automatically recognizing and describing images in text, images can be converted into text and then into speech for visually impaired people to help them understand their surroundings, and important issues such as image search, art therapy, sports commentary, and real-time traffic information commentary. So far, the image captioning research approach focuses solely on recognizing and texturing images. However, various environments in reality must be considered for practical use, as well as being able to provide image descriptions for the intended purpose. In this work, we limit the universally available Korean and English image captioning models and text generation techniques for the purpose of image captioning.

• 한국어정보학회:학술대회논문집
• /
• 2016.10a
• /
• pp.176-180
• /
• 2016

대화형 시스템이 사람의 경청 기술을 모방할 수 있다면 대화 상대방과 더 효과적으로 상호작용 할 수 있을 것이다. 본 논문에서는 시스템이 경청 기술을 모방할 수 있도록 사용자의 발화를 기반으로 질문을 생성하는 것에 대해 연구하였다. 그리고 이러한 연구를 위해 필요한 데이터를 Image captioning과 Visual QA 데이터를 기반으로 생성하고 활용하는 방안에 대해 제안한다. 또한 이러한 데이터를 Attention 메커니즘을 적용한 Sequence to sequence 모델에 적용하여 질문을 생성하고, 생성된 질문의 질문 유형을 분석하였다. 마지막으로 사람이 작성한 질문과 모델의 질문 생성 결과 비교를 BLEU 점수를 이용하여 수행하였다.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.176-180
• /
• 2016

대화형 시스템이 사람의 경청 기술을 모방할 수 있다면 대화 상대방과 더 효과적으로 상호작용 할 수 있을 것이다. 본 논문에서는 시스템이 경청 기술을 모방할 수 있도록 사용자의 발화를 기반으로 질문을 생성하는 것에 대해 연구하였다. 그리고 이러한 연구를 위해 필요한 데이터를 Image captioning과 Visual QA 데이터를 기반으로 생성하고 활용하는 방안에 대해 제안한다. 또한 이러한 데이터를 Attention 메커니즘을 적용한 Sequence to sequence 모델에 적용하여 질문을 생성하고, 생성된 질문의 질문 유형을 분석하였다. 마지막으로 사람이 작성한 질문과 모델의 질문 생성 결과 비교를 BLEU 점수를 이용하여 수행하였다.

• ETRI Journal
• /
• v.42 no.1
• /
• pp.67-77
• /
• 2020

Image captioning has received significant interest in recent years, and notable results have been achieved. Most previous approaches have focused on generating visual descriptions from images, whereas a few approaches have exploited visual descriptions for image classification. This study demonstrates that a good performance can be achieved for both description generation and image classification through an end-to-end joint learning approach with a loss function, which encourages each task to reach a consensus. When given images and visual descriptions, the proposed model learns a multimodal intermediate embedding, which can represent both the textual and visual characteristics of an object. The performance can be improved for both tasks by sharing the multimodal embedding. Through a novel loss function based on class activation mapping, which localizes the discriminative image region of a model, we achieve a higher score when the captioning and classification model reaches a consensus on the key parts of the object. Using the proposed model, we established a substantially improved performance for each task on the UCSD Birds and Oxford Flowers datasets.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.01a
• /
• pp.303-305
• /
• 2023

본 논문에서는 트랜스포머를 사용한 이미지 캡셔닝 방법과 비디오 캡셔닝 방법을 제안한다. 트랜스포머의 입력으로 사전 학습된 이미지 클래스 분류모델을 거쳐 추출된 특징을 트랜스포머의 입력으로 넣고 인코더-디코더를 통해 이미지와 비디오의 캡션을 출력한다. 이미지 캡셔닝의 경우 한글 데이터 세트를 학습하여 한글 캡션을 출력하도록 학습하였으며 비디오 캡셔닝의 경우 MSVD 데이터 세트를 학습하여 학습 후 출력 캡션의 성능을 다른 비디오 캡셔닝 모델의 성능과 비교하였다. 비디오 캡셔닝에서 성능향상을 위해 트랜스포머의 디코더를 변형한 GPT-2를 사용하였을 때 BLEU-1 점수가 트랜스포머의 경우 0.62, GPT-2의 경우 0.80으로 성능이 향상됨을 확인하였다

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.10
• /
• pp.3390-3405
• /
• 2022

Long Short-Term Memory (LSTM) combined with attention mechanism is extensively used to generate semantic sentences of images in image captioning models. However, features of salient regions and spatial information are not utilized sufficiently in most related works. Meanwhile, the LSTM also suffers from the problem of underutilized information in a single time step. In the paper, two innovative approaches are proposed to solve these problems. First, the Synergy-Gated Attention (SGA) method is proposed, which can process the spatial features and the salient region features of given images simultaneously. SGA establishes a gated mechanism through the global features to guide the interaction of information between these two features. Then, the Recurrent Fusion LSTM (RF-LSTM) mechanism is proposed, which can predict the next hidden vectors in one time step and improve linguistic coherence by fusing future information. Experimental results on the benchmark dataset of MSCOCO show that compared with the state-of-the-art methods, the proposed method can improve the performance of image captioning model, and achieve competitive performance on multiple evaluation indicators.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.79-104
• /
• 2020

Recently, as deep learning has attracted attention, the use of deep learning is being considered as a method for solving problems in various fields. In particular, deep learning is known to have excellent performance when applied to applying unstructured data such as text, sound and images, and many studies have proven its effectiveness. Owing to the remarkable development of text and image deep learning technology, interests in image captioning technology and its application is rapidly increasing. Image captioning is a technique that automatically generates relevant captions for a given image by handling both image comprehension and text generation simultaneously. In spite of the high entry barrier of image captioning that analysts should be able to process both image and text data, image captioning has established itself as one of the key fields in the A.I. research owing to its various applicability. In addition, many researches have been conducted to improve the performance of image captioning in various aspects. Recent researches attempt to create advanced captions that can not only describe an image accurately, but also convey the information contained in the image more sophisticatedly. Despite many recent efforts to improve the performance of image captioning, it is difficult to find any researches to interpret images from the perspective of domain experts in each field not from the perspective of the general public. Even for the same image, the part of interests may differ according to the professional field of the person who has encountered the image. Moreover, the way of interpreting and expressing the image also differs according to the level of expertise. The public tends to recognize the image from a holistic and general perspective, that is, from the perspective of identifying the image's constituent objects and their relationships. On the contrary, the domain experts tend to recognize the image by focusing on some specific elements necessary to interpret the given image based on their expertise. It implies that meaningful parts of an image are mutually different depending on viewers' perspective even for the same image. So, image captioning needs to implement this phenomenon. Therefore, in this study, we propose a method to generate captions specialized in each domain for the image by utilizing the expertise of experts in the corresponding domain. Specifically, after performing pre-training on a large amount of general data, the expertise in the field is transplanted through transfer-learning with a small amount of expertise data. However, simple adaption of transfer learning using expertise data may invoke another type of problems. Simultaneous learning with captions of various characteristics may invoke so-called 'inter-observation interference' problem, which make it difficult to perform pure learning of each characteristic point of view. For learning with vast amount of data, most of this interference is self-purified and has little impact on learning results. On the contrary, in the case of fine-tuning where learning is performed on a small amount of data, the impact of such interference on learning can be relatively large. To solve this problem, therefore, we propose a novel 'Character-Independent Transfer-learning' that performs transfer learning independently for each character. In order to confirm the feasibility of the proposed methodology, we performed experiments utilizing the results of pre-training on MSCOCO dataset which is comprised of 120,000 images and about 600,000 general captions. Additionally, according to the advice of an art therapist, about 300 pairs of 'image / expertise captions' were created, and the data was used for the experiments of expertise transplantation. As a result of the experiment, it was confirmed that the caption generated according to the proposed methodology generates captions from the perspective of implanted expertise whereas the caption generated through learning on general data contains a number of contents irrelevant to expertise interpretation. In this paper, we propose a novel approach of specialized image interpretation. To achieve this goal, we present a method to use transfer learning and generate captions specialized in the specific domain. In the future, by applying the proposed methodology to expertise transplant in various fields, we expected that many researches will be actively conducted to solve the problem of lack of expertise data and to improve performance of image captioning.

• The Journal of Information Systems
• /
• v.32 no.4
• /
• pp.119-136
• /
• 2023

Purpose We introduce a groundbreaking surveillance system explicitly designed to overcome the limitations typically associated with conventional surveillance systems, which often focus primarily on object-centric behavior analysis. Design/methodology/approach The study introduces an innovative approach to risk assessment in surveillance, employing image captioning to generate descriptive captions that effectively encapsulate the interactions among objects, actions, and spatial elements within observed scenes. To support our methodology, we developed a distinctive dataset comprising pairs of [image-caption-danger score] for training purposes. We fine-tuned the BLIP-2 model using this dataset and utilized BERT to decipher the semantic content of the generated captions for assessing risk levels. Findings In a series of experiments conducted with our self-constructed datasets, we illustrate that these datasets offer a wealth of information for risk assessment and display outstanding performance in this area. In comparison to models pre-trained on established datasets, our generated captions thoroughly encompass the necessary object attributes, behaviors, and spatial context crucial for the surveillance system. Additionally, they showcase adaptability to novel sentence structures, ensuring their versatility across a range of contexts.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.101-125
• /
• 2022

Recently, with the development of computing technology and the improvement of the cloud environment, deep learning technology has developed, and attempts to apply deep learning to various fields are increasing. A typical example is anomaly detection, which is a technique for identifying values or patterns that deviate from normal data. Among the representative types of anomaly detection, it is very difficult to detect a contextual anomaly that requires understanding of the overall situation. In general, detection of anomalies in image data is performed using a pre-trained model trained on large data. However, since this pre-trained model was created by focusing on object classification of images, there is a limit to be applied to anomaly detection that needs to understand complex situations created by various objects. Therefore, in this study, we newly propose a two-step pre-trained model for detecting abnormal situation. Our methodology performs additional learning from image captioning to understand not only mere objects but also the complicated situation created by them. Specifically, the proposed methodology transfers knowledge of the pre-trained model that has learned object classification with ImageNet data to the image captioning model, and uses the caption that describes the situation represented by the image. Afterwards, the weight obtained by learning the situational characteristics through images and captions is extracted and fine-tuning is performed to generate an anomaly detection model. To evaluate the performance of the proposed methodology, an anomaly detection experiment was performed on 400 situational images and the experimental results showed that the proposed methodology was superior in terms of anomaly detection accuracy and F1-score compared to the existing traditional pre-trained model.

• Journal of KIISE:Software and Applications
• /
• v.31 no.10
• /
• pp.1386-1399
• /
• 2004

This paper presents correlation-based automatic image captioning. Given a training set of annotated images, we want to discover correlations between visual features and textual features, so that we can automatically generate descriptive textual features for a new unseen image. We develop models with multiple design alternatives such as 1) adaptively clustering visual features, 2) weighting visual features and textual features, and 3) reducing dimensionality for noise sup-Pression. We experiment thoroughly on 10 data sets of various content styles from the Corel image database, about 680MB. The major contributions of this work are: (a) we show that careful weighting visual and textual features, as well as clustering visual features adaptively leads to consistent performance improvements, and (b) our proposed methods achieve a relative improvement of up to 45% on annotation accuracy over the state-of-the-art, EM approach.