Daw-Tung Lin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Daw-Tung Lin has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Bor-Chen Tsai
CCP Contact Probes CO
1 shared publication - Chih-Feng Chen
CCP Contact Probes CO
1 shared publication Chin-Hui Hsu
National Taipei University
1 shared publication- Chun-Seng Chiao
Department of Computer Science and Information Engineering, National Taipei University, 151 University Road, San-Xia Dist., New Taipei City, Taiwan
1 shared publication Chung-Chih Lei
1 shared publicationChung-Han Wu
Inst. of Electr. Eng., Nat. Taipei Univ., Taipei, Taiwan
1 shared publication
Latest Publications
- 25 April 2026
- Published by Elsevier in Biomedical Signal Processing and Control
- Vol. 122
Deep learning-based screening for mild cognitive impairment through clock-drawing test
- 25 April 2026
- Published by Elsevier in Biomedical Signal Processing and Control
- Vol. 122
Deep learning-based screening for mild cognitive impairment through clock-drawing test
- 17 September 2025
- Published by Springer Nature
- p. 39-52
A Region-Aware Multi-modal Framework for Pedestrian Attribute Recognition via CLIP and Graph Neural Networks
- 17 September 2025
- Published by Springer Nature
- p. 39-52
A Region-Aware Multi-modal Framework for Pedestrian Attribute Recognition via CLIP and Graph Neural Networks
- 29 May 2025
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (36) , 45095-45117
Enhancing multi-target multi-camera vehicle tracking with YOLOv9 and attention mechanisms for smart city traffic monitoring
- 29 May 2025
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (36) , 45095-45117
Enhancing multi-target multi-camera vehicle tracking with YOLOv9 and attention mechanisms for smart city traffic monitoring
- 27 May 2025
- Published by Springer Nature
- p. 93-105
Advancing Image Registration with Multi-angle Projection Keypoint Descriptors
- 27 May 2025
- Published by Springer Nature
- p. 93-105
Advancing Image Registration with Multi-angle Projection Keypoint Descriptors
- 21 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (1) , 4425-4434
System Implementation of Multiple License Plate Detection and Correction on Wide-Angle Images Using an Instance Segmentation Network Model
The shift from traditional image processing to advanced deep learning has greatly improved license plate recognition, increasing its maturity and stability. Despite these advancements, challenges persist, including susceptibility to environmental factors like insufficient lighting, complex backgrounds, and low-resolution images. Detection of multiple license plates involves processing various targets characterized by distinct sizes, lighting conditions, occlusion, positions, and angles. Each license plate may be located at different positions and orientations within the image, significantly increasing the complexity of the detection process. The study, conducted in an indoor parking lot with a wide-angle camera, addresses perspective-warping uncertainties. Utilizing the Segmenting Objects by Locations version 2 (SOLOv2) SOLOv2 model enhances multiple license plate detection and segmentation, with the resulting mask aiding corner location and correction, leading to a 12% recognition rate increase. Notably, the study employs an instance segmentation network, enabling simultaneous recognition of multiple vehicle license plates, offering a significant advancement in overcoming complexities associated with diverse environmental conditions and camera angles.
- 21 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (1) , 4425-4434
System Implementation of Multiple License Plate Detection and Correction on Wide-Angle Images Using an Instance Segmentation Network Model
The shift from traditional image processing to advanced deep learning has greatly improved license plate recognition, increasing its maturity and stability. Despite these advancements, challenges persist, including susceptibility to environmental factors like insufficient lighting, complex backgrounds, and low-resolution images. Detection of multiple license plates involves processing various targets characterized by distinct sizes, lighting conditions, occlusion, positions, and angles. Each license plate may be located at different positions and orientations within the image, significantly increasing the complexity of the detection process. The study, conducted in an indoor parking lot with a wide-angle camera, addresses perspective-warping uncertainties. Utilizing the Segmenting Objects by Locations version 2 (SOLOv2) SOLOv2 model enhances multiple license plate detection and segmentation, with the resulting mask aiding corner location and correction, leading to a 12% recognition rate increase. Notably, the study employs an instance segmentation network, enabling simultaneous recognition of multiple vehicle license plates, offering a significant advancement in overcoming complexities associated with diverse environmental conditions and camera angles.
- 25 May 2023
- Published by Elsevier in Expert Systems with Applications
- Vol. 229
ELICE: Embedding Language through Informative Contrastive-Encoder
- 25 May 2023
- Published by Elsevier in Expert Systems with Applications
- Vol. 229
ELICE: Embedding Language through Informative Contrastive-Encoder
- Open Access
- 17 January 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (2) , 907
Large-Scale Printed Chinese Character Recognition for ID Cards Using Deep Learning and Few Samples Transfer Learning
In the field of computer vision, large-scale image classification tasks are both important and highly challenging. With the ongoing advances in deep learning and optical character recognition (OCR) technologies, neural networks designed to perform large-scale classification play an essential role in facilitating OCR systems. In this study, we developed an automatic OCR system designed to identify up to 13,070 large-scale printed Chinese characters by using deep learning neural networks and fine-tuning techniques. The proposed framework comprises four components, including training dataset synthesis and background simulation, image preprocessing and data augmentation, the process of training the model, and transfer learning. The training data synthesis procedure is composed of a character font generation step and a background simulation process. Three background models are proposed to simulate the factors of the background noise patterns on ID cards. To expand the diversity of the synthesized training dataset, rotation and zooming data augmentation are applied. A massive dataset comprising more than 19.6 million images was thus created to accommodate the variations in the input images and improve the learning capacity of the CNN model. Subsequently, we modified the GoogLeNet neural architecture by replacing the fully connected layer with a global average pooling layer to avoid overfitting caused by a massive amount of training data. Consequently, the number of model parameters was reduced. Finally, we employed the transfer learning technique to further refine the CNN model using a small number of real data samples. Experimental results show that the overall recognition performance of the proposed approach is significantly better than that of prior methods and thus demonstrate the effectiveness of proposed framework, which exhibited a recognition accuracy as high as 99.39% on the constructed real ID card dataset.
- 17 January 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (2) , 907
Open AccessLarge-Scale Printed Chinese Character Recognition for ID Cards Using Deep Learning and Few Samples Transfer Learning
In the field of computer vision, large-scale image classification tasks are both important and highly challenging. With the ongoing advances in deep learning and optical character recognition (OCR) technologies, neural networks designed to perform large-scale classification play an essential role in facilitating OCR systems. In this study, we developed an automatic OCR system designed to identify up to 13,070 large-scale printed Chinese characters by using deep learning neural networks and fine-tuning techniques. The proposed framework comprises four components, including training dataset synthesis and background simulation, image preprocessing and data augmentation, the process of training the model, and transfer learning. The training data synthesis procedure is composed of a character font generation step and a background simulation process. Three background models are proposed to simulate the factors of the background noise patterns on ID cards. To expand the diversity of the synthesized training dataset, rotation and zooming data augmentation are applied. A massive dataset comprising more than 19.6 million images was thus created to accommodate the variations in the input images and improve the learning capacity of the CNN model. Subsequently, we modified the GoogLeNet neural architecture by replacing the fully connected layer with a global average pooling layer to avoid overfitting caused by a massive amount of training data. Consequently, the number of model parameters was reduced. Finally, we employed the transfer learning technique to further refine the CNN model using a small number of real data samples. Experimental results show that the overall recognition performance of the proposed approach is significantly better than that of prior methods and thus demonstrate the effectiveness of proposed framework, which exhibited a recognition accuracy as high as 99.39% on the constructed real ID card dataset.
- 7 May 2021
- Published by Springer Nature in Journal of Digital Imaging
- Vol. 34 (3) , 637-646
Rapid Assessment of Acute Ischemic Stroke by Computed Tomography Using Deep Convolutional Neural Networks
Acute stroke is one of the leading causes of disability and death worldwide. Regarding clinical diagnoses, a rapid and accurate procedure is necessary for patients suffering from acute stroke. This study proposes an automatic identification scheme for acute ischemic stroke using deep convolutional neural networks (DCNNs) based on non-contrast computed tomographic (NCCT) images. Our image database for the classification model was composed of 1254 grayscale NCCT images from 96 patients (573 images) with acute ischemic stroke and 121 normal controls (681 images). According to the consensus of critical stroke findings by two neuroradiologists, a gold standard was established and used to train the proposed DCNN using machine-generated image features. Including the earliest DCNN, AlexNet, the popular Inception-v3, and ResNet-101 were proposed. To train the limited data size, transfer learning with ImageNet parameters was also used. The established models were evaluated by tenfold cross-validation and tested on an independent dataset containing 50 patients with acute ischemic stroke (108 images) and 58 normal controls (117 images) from another institution. AlexNet without pretrained parameters achieved an accuracy of 97.12%, a sensitivity of 98.11%, a specificity of 96.08%, and an area under the receiver operating characteristic curve (AUC) of 0.9927. Using transfer learning, transferred AlexNet, transferred Inception-v3, and transferred ResNet-101 achieved accuracies between 90.49 and 95.49%. Tested with a dataset from another institution, AlexNet showed an accuracy of 60.89%, a sensitivity of 18.52%, and a specificity of 100%. Transferred AlexNet, Inception-v3, and ResNet-101 achieved accuracies of 81.77%, 85.78%, and 80.89%, respectively. The proposed DCNN architecture as a computer-aided diagnosis system showed that training from scratch can generate a customized model for a specific scanner, and transfer learning can generate a more generalized model to provide diagnostic suggestions of acute ischemic stroke to radiologists.
- 7 May 2021
- Published by Springer Nature in Journal of Digital Imaging
- Vol. 34 (3) , 637-646
Rapid Assessment of Acute Ischemic Stroke by Computed Tomography Using Deep Convolutional Neural Networks
Acute stroke is one of the leading causes of disability and death worldwide. Regarding clinical diagnoses, a rapid and accurate procedure is necessary for patients suffering from acute stroke. This study proposes an automatic identification scheme for acute ischemic stroke using deep convolutional neural networks (DCNNs) based on non-contrast computed tomographic (NCCT) images. Our image database for the classification model was composed of 1254 grayscale NCCT images from 96 patients (573 images) with acute ischemic stroke and 121 normal controls (681 images). According to the consensus of critical stroke findings by two neuroradiologists, a gold standard was established and used to train the proposed DCNN using machine-generated image features. Including the earliest DCNN, AlexNet, the popular Inception-v3, and ResNet-101 were proposed. To train the limited data size, transfer learning with ImageNet parameters was also used. The established models were evaluated by tenfold cross-validation and tested on an independent dataset containing 50 patients with acute ischemic stroke (108 images) and 58 normal controls (117 images) from another institution. AlexNet without pretrained parameters achieved an accuracy of 97.12%, a sensitivity of 98.11%, a specificity of 96.08%, and an area under the receiver operating characteristic curve (AUC) of 0.9927. Using transfer learning, transferred AlexNet, transferred Inception-v3, and transferred ResNet-101 achieved accuracies between 90.49 and 95.49%. Tested with a dataset from another institution, AlexNet showed an accuracy of 60.89%, a sensitivity of 18.52%, and a specificity of 100%. Transferred AlexNet, Inception-v3, and ResNet-101 achieved accuracies of 81.77%, 85.78%, and 80.89%, respectively. The proposed DCNN architecture as a computer-aided diagnosis system showed that training from scratch can generate a customized model for a specific scanner, and transfer learning can generate a more generalized model to provide diagnostic suggestions of acute ischemic stroke to radiologists.
- 6 February 2020
- Published by Springer Nature
- p. 360-369
Automatic Optical Inspection for Millimeter Scale Probe Surface Stripping Defects Using Convolutional Neural Network
- 6 February 2020
- Published by Springer Nature
- p. 360-369
Automatic Optical Inspection for Millimeter Scale Probe Surface Stripping Defects Using Convolutional Neural Network
- 9 February 2019
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 79 (15-16) , 11125-11141
Effective multiple person recognition in random video sequences using a convolutional neural network
- 9 February 2019
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 79 (15-16) , 11125-11141
Effective multiple person recognition in random video sequences using a convolutional neural network