Article
Chemistry, Multidisciplinary
Ngoc-Hoang Nguyen, Tran-Dac-Thinh Phan, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee
Summary: This paper introduces a novel approach to continuous dynamic hand gesture recognition, utilizing two main modules: gesture spotting and gesture classification. By combining different data channels and conducting experiments on multiple public datasets, the approach shows promising performance improvements in gesture recognition.
APPLIED SCIENCES-BASEL
(2021)
Article
Engineering, Civil
Gibran Benitez-Garcia, Muhammad Haris, Yoshiyuki Tsuda, Norimichi Ukita
Summary: This article introduces a touchless technology for controlling in-car devices using finger gestures. The proposed method achieves real-time performance by segmenting and recognizing gestures, and a dataset of continuous finger gestures is provided for model validation.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Robotics
Marc Peral, Alberto Sanfeliu, Anais Garrell
Summary: In this paper, an efficient and reliable deep-learning approach for hand gesture recognition in robot communication is presented. The approach utilizes visual information to extract hand landmarks and predict gestures, achieving high accuracy and real-time performance.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Computer Science, Hardware & Architecture
Graziano Fronteddu, Simone Porcu, Alessandro Floris, Luigi Atzori
Summary: In this article, a dataset consisting of 27 dynamic hand gesture types is proposed. The dataset was collected from 21 subjects at full HD resolution. The subjects were carefully instructed and monitored to ensure the accuracy of the gestures.
Article
Computer Science, Artificial Intelligence
Anish K. Monsley, Kuldeep Singh Yadav, Songhita Misra, Rabul Hussain Laskar, Taimoor Khan, M. K. Bhuyan
Summary: The study introduces a gesture recognition method based on an artificial neural network, which preprocesses and classifies gestures to reduce the impact of self-coarticulation, thereby improving the accuracy of gesture recognition.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Information Systems
Hong-Quan Nguyen, Trung-Hieu Le, Trung-Kien Tran, Hoang-Nhat Tran, Thanh-Hai Tran, Thi-Lan Le, Hai Vu, Cuong Pham, Thanh Phuong Nguyen, Huu Thanh Nguyen
Summary: In this work, we explore the use of hand gestures for human-machine interaction using wrist-worn sensors. We develop a prototype that captures RGB video stream of hand gestures and create a new gesture dataset. We evaluate various CNN models and find that MoviNet produces the highest accuracy. Additionally, we propose a two-stream architecture based on MoviNet that incorporates both RGB and optical flow, leading to improved recognition accuracy.
Article
Computer Science, Artificial Intelligence
Mohammad Mahmudul Alam, Mohammad Tariqul Islam, S. M. Mahbubur Rahman
Summary: This paper introduces a unified approach for egocentric hand gesture recognition and fingertip detection, using a single convolutional neural network to predict the probabilities of finger class and positions of fingertips. The proposed method shows improved computation speed and outperforms existing approaches in experimental results.
PATTERN RECOGNITION
(2022)
Article
Chemistry, Analytical
Lianqing Zheng, Jie Bai, Xichan Zhu, Libo Huang, Chewu Shan, Qiong Wu, Lei Zhang
Summary: This paper proposes a gesture recognition system based on radar and transformer for in-vehicle environment, and designs a transformer-based radar gesture recognition network named RGTNet. Experimental results show that this method can better complete gesture classification tasks in the in-vehicle environment with an accuracy of 97.56%.
Article
Computer Science, Information Systems
Jun Xu, Hanchen Wang, Jianrong Zhang, Linqin Cai
Summary: This paper presents a robust RGB-D data-based recognition method for static and dynamic hand gestures, utilizing algorithms like Distance Transform and K-Curvature-Convex Defects Detection for gesture identification and feature vector construction, and proposing recognition algorithms. Additionally, a unifying feature descriptor is generated for dynamic gestures by combining Euclidean distance and skeleton feature ratios for recognition. Extensive experiments validate the real-time application of the gesture recognition algorithm.
Article
Computer Science, Artificial Intelligence
Qiyu Li, Reza Langari
Summary: Human-computer interaction (HCI) has diverse applications, with rehabilitation devices being one important domain. Surface ElectroMyoGraphic (sEMG) signals are commonly used for myoelectric control, but challenges remain in accurately recognizing gestures produced by hand or upper arm due to limb positions. This paper proposes a CNN-LSTM model for dynamic gesture recognition, achieving an overall accuracy of 84.2% with potential for up to 90% accuracy for some subjects.
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2023)
Review
Chemistry, Multidisciplinary
Zhihan Lv, Fabio Poiesi, Qi Dong, Jaime Lloret, Houbing Song
Summary: Gesture recognition and speech recognition, as important input methods in Human-Computer Interaction (HCI), have achieved breakthrough research progress with the rapid development of deep learning and artificial intelligence. This study analyzes the current situation of HCI intelligent systems and summarizes the implementation of gesture interaction and voice interaction in HCI. The combination of intelligent HCI and deep learning is deeply applied in gesture recognition, speech recognition, emotion recognition, and intelligent robot direction, resulting in high recognition accuracy and better robustness in Human-Machine Interfaces (HMIs) with voice support.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Interdisciplinary Applications
Damon Shing-Min Liu, Shao-Jun Wu
Summary: This paper discusses the importance of improving the realism of augmented reality by focusing on lighting consistency and hand interaction. The authors propose a method to estimate light direction based on shadows and foreground objects, and introduce gesture recognition and hand touchable interaction to enhance the realism of augmented reality.
Article
Engineering, Electrical & Electronic
Qiang Fu, Jiajun Fu, Songyuan Zhang, Xun Li, Jian Guo, Shuxiang Guo
Summary: The study designed an intelligent human-computer interaction system that effectively addresses communication inconvenience between the hard of hearing and the non-disabled by combining artificial intelligence with wearable devices and classifying gestures using a BP neural network.
IEEE SENSORS JOURNAL
(2021)
Article
Computer Science, Artificial Intelligence
Sakshi Sharma, Sukhwinder Singh
Summary: Hand gestures are crucial for communication and form the foundation of sign language, which is a visual form of communication. A deep learning CNN model designed for recognizing gesture-based sign language achieved high classification accuracy with fewer model parameters. The proposed model outperformed existing techniques in classifying gestures accurately with minimal error rates.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Automation & Control Systems
Chen Zhu, Jianyu Yang, Zhanpeng Shao, Chunping Liu
Summary: The paper introduces a new method for hand gesture recognition using depth maps and 3D shape context descriptors. Experimental results show that the proposed method is robust to noise, articulated variations, and rigid transformations, outperforming current state-of-the-art methods in accuracy and efficiency.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2021)
Article
Computer Science, Information Systems
Dinh-Son Tran, Ngoc-Huynh Ho, Hyung-Jeong Yang, Soo-Hyung Kim, Guee Sang Lee
Summary: The study introduces a novel virtual-mouse method using RGB-D images and fingertip detection, which can track fingertip location in real-time and operate efficiently in real-world environments, enabling easy interaction with computers through gestures.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Chemistry, Multidisciplinary
Minh-Trieu Tran, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee
Summary: The restoration of medical images is crucial for accurate diagnosis, and the multi-task learning model proposed in this paper achieves innovative results in medical image inpainting, generating more realistic and believable images. The method shows excellent performance in quantitative evaluation and brings new possibilities to the field of medical image analysis.
APPLIED SCIENCES-BASEL
(2021)
Article
Medicine, General & Internal
Nhu-Tai Do, Sung-Taek Jung, Hyung-Jeong Yang, Soo-Hyung Kim
Summary: This study aims to classify and segment knee bone tumors using deep learning, proposing the Seg-Unet model and building a corresponding dataset, with experiments showing superior performance in accuracy.
Article
Chemistry, Analytical
Tran-Dac-Thinh Phan, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee
Summary: This study focuses on emotion recognition based on EEG signals, utilizing time domain features and the correlation between 32 channels and frequency bands to enhance prediction performance. Using a 2D CNN structure to process EEG signals enables learning of relationships between local and global patterns of channels.
Article
Chemistry, Analytical
Minh-Trieu Tran, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee, In-Jae Oh, Sae-Ryung Kang
Summary: This study proposes a fully automated framework for esophagus segmentation, addressing the challenges of its small size, ambiguous boundary, and low contrast in CT images. By utilizing spatial attention mechanism and advanced modules, the method effectively locates the esophagus and enhances segmentation performance. Optimization using the STAPLE algorithm improves Dice and Hausdorff Distance scores for segmentation results.
Article
Chemistry, Multidisciplinary
Ngoc-Hoang Nguyen, Tran-Dac-Thinh Phan, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee
Summary: This paper introduces a novel approach to continuous dynamic hand gesture recognition, utilizing two main modules: gesture spotting and gesture classification. By combining different data channels and conducting experiments on multiple public datasets, the approach shows promising performance improvements in gesture recognition.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Artificial Intelligence
Ngoc-Huynh Ho, Hyung-Jeong Yang, Jahae Kim, Duy-Phuong Dao, Hyuk-Ro Park, Sudarshan Pant
Summary: This study proposes a progressive recurrent network model for predicting clinical diagnoses and phenotypic measurements of Alzheimer's disease patients using longitudinal data. The model utilizes magnetic resonance imaging data and an imputation module to compensate for missing observations in the data.
Article
Chemistry, Multidisciplinary
Eu-Tteum Baek, Dae-Yeong Im
Summary: The paper proposes a method to design and operate a mobile robot platform for use in a greenhouse, which is expected to increase productivity and reduce labor costs.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Information Systems
Dang-Linh Trinh, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee
Summary: Glioblastoma is a highly aggressive brain malignancy, and accurate survival prediction is crucial for diagnosis and treatment planning. This study demonstrates the effectiveness of features extracted from whole tumor and enhancing tumor in predicting overall survival. Optimal shape radiomics features and deep features are selected and combined with clinical information to build a regression model. The proposed method achieves promising results in accuracy and mean squared error metrics.
Article
Chemistry, Analytical
Dang-Linh Trinh, Minh-Cong Vo, Soo-Hyung Kim, Hyung-Jeong Yang, Guee-Sang Lee
Summary: This paper proposes a self-relation attention and temporal awareness (SRA-TA) module to tackle the problem of emotion recognition via non-verbal speech. The proposed method achieves a high performance on the test set, ranking first in the 2022 ACII Affective Vocal Burst Workshop & Challenge.
Article
Chemistry, Analytical
Irfan Haider, Hyung-Jeong Yang, Guee-Sang Lee, Soo-Hyung Kim
Summary: Human facial emotion detection is a challenging task in computer vision. This paper proposes a novel and intelligent approach for emotion classification. It uses a customized ResNet18 model with transfer learning and triplet loss function, combined with an SVM classifier to accurately predict facial emotions. Experimental results show that the proposed method achieves high accuracy on JAFFE and MMI datasets.
Article
Health Care Sciences & Services
Sudarshan Pant, Sae-Ryung Kang, Minhee Lee, Pham-Sy Phuc, Hyung-Jeong Yang, Deok-Hwan Yang
Summary: This study aims to develop a robust survival prediction strategy for diffuse large B-cell lymphoma (DLBCL) using a deep-learning-based approach. The proposed model combines clinical risk factors and Deauville scores in positron-emission tomography/computed tomography at different treatment stages. Results show that the model outperforms existing methods in terms of survival time estimation. The use of Deauville scores during treatment improves prognostic accuracy.
Article
Computer Science, Information Systems
Thi-Dung Tran, Ngoc-Huynh Ho, Sudarshan Pant, Hyung-Jeong Yang, Soo-Hyung Kim, Gueesang Lee
Summary: Humans can determine subtle emotions from various indicators and surroundings. However, existing research on emotion recognition mainly focuses on recognizing the emotions of speakers. Thus, this paper proposes a novel multimodal approach for predicting emotions from missing modalities.
Review
Health Care Sciences & Services
Ngumimi Karen Iyortsuun, Soo-Hyung Kim, Min Jhon, Hyung-Jeong Yang, Sudarshan Pant
Summary: Machine learning approaches have been used in the diagnosis and prediction of treatment outcomes for mental health conditions. This study reviewed 33 articles on the use of machine learning and deep learning technologies for the diagnosis of various mental disorders, including schizophrenia, depression, anxiety, bipolar disorder, PTSD, anorexia nervosa, and ADHD. The researchers also discussed the challenges they encountered and provided a list of public datasets.
Proceedings Paper
Computer Science, Artificial Intelligence
Duy-Phuong Dao, Hyung-Jeong Yang, Ngoc-Huynh Ho, Sudarshan Pant, Soo-Hyung Kim, Guee-Sang Lee, In-Jae Oh, Sae-Ryung Kang
Summary: Lung cancer is the most common type of cancer, and its incidence among young people is increasing. Early diagnosis and treatment are crucial for successful recovery. This study proposes a tumor segmentation model called MAPTransNet and a survival analysis model called MSNet to assist in the diagnosis and prediction of lung cancer patients.
2022 26TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR)
(2022)
