Mei-Hua Hsih

Dr. Mei-Hua Hsih


Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
20 April 2024 in Information
Verified
Journal Article
Open access
PDF file
Chen-Chiung HsiehChen-Chiung Hsieh
Chen-Chiung Hsieh
HJHan-Wen Jia
Han-Wen Jia
WHWei-Hsin Huang
Wei-Hsin Huang
Mei-Hua HsihMei-Hua Hsih
Mei-Hua Hsih
This study proposes a deep learning method for pavement defect detection, focusing on identifying potholes and cracks. A dataset comprising 10,828 images is collected, with 8662 allocated for training, 1083 for validation, and 1083 for testing. Vehicle attitude data are categorized based on three-axis acceleration and attitude change, with 6656 (64%) for training, 1664 (16%) for validation, and 2080 (20%) for testing. The Nvidia Jetson Nano serves as the vehicle-embedded system, transmitting IMU-acquired vehicle data and GoPro-captured images over a 5G network to the server. The server recognizes two damage categories, low-risk and high-risk, storing results in MongoDB. Severe damage triggers immediate alerts to maintenance personnel, while less severe issues are recorded for scheduled maintenance. The method selects YOLOv7 among various object detection models for pavement defect detection, achieving a mAP of 93.3%, a recall rate of 87.8%, a precision of 93.2%, and a processing speed of 30–40 FPS. Bi-LSTM is then chosen for vehicle vibration data processing, yielding 77% mAP, 94.9% recall rate, and 89.8% precision. Integration of the visual and vibration results, along with vehicle speed and travel distance, results in a final recall rate of 90.2% and precision of 83.7% after field testing.

21 October 2025 in Electronics
Verified
Journal Article
Open access
PDF file
Mei-Hua HsihMei-Hua Hsih
Mei-Hua Hsih
SLShih-Po Lin
Shih-Po Lin
Chen-Chiung HsiehChen-Chiung Hsieh
Chen-Chiung Hsieh
Medical imaging, particularly chest X-rays, plays a vital role in radiological diagnosis. However, interpreting these images and generating detailed diagnostic reports is a time-consuming task for clinicians. To address this challenge, this study proposes an automated image captioning framework for chest X-ray images, aiming to reduce clinical workload and enhance diagnostic efficiency. The proposed approach employs convolutional neural networks (CNNs) for visual feature extraction and a modified Transformer architecture—referred to as the Medical Transformer—for structured report generation. Three CNN models, namely InceptionV3, ResNet152V2, and Inception–ResNetV2, were evaluated as feature extractors. The attention mechanisms, Bahdanau, Luong, and scaled dot product, were activated by ReLU or Tanh functions to identify the optimal configuration, i.e., the maximum attention is used. Experiments were conducted using the Indiana University Chest X-ray dataset, which contains 7466 images paired with corresponding diagnostic reports. The proposed approach employs image augmentation to accommodate input variability, utilizes Inception–ResNetV2 for feature extraction, and integrates the Medical Transformer with maximum attention mechanisms to achieve optimal performance in medical report generation. Evaluation metrics include BLEU (BLEU-1 to BLEU-4 scores of 0.720, 0.669, 0.648, and 0.600, respectively), METEOR (0.741), and BERTScore (FBERT = 0.787), demonstrating superior performance compared to baseline models and the state of the art. These results validate the effectiveness of the proposed Medical Transformer framework in generating accurate and clinically relevant medical image captions.

20 December 2024 in Information
Verified
Journal Article
Open access
PDF file
Mei-Hua HsihMei-Hua Hsih
Mei-Hua Hsih
JYJian-Xin Yang
Jian-Xin Yang
Chen-Chiung HsiehChen-Chiung Hsieh
Chen-Chiung Hsieh
This paper proposes using a web crawler to organize website content as a dialogue tree in some domains. We build an intelligent customer service agent based on this dialogue tree for general usage. The encoder-decoder architecture Seq2Seq is used to understand natural language and then modified as a bi-directional LSTM to increase the accuracy of the polysemy cases. The attention mechanism is added in the decoder to improve the problem of accuracy decreasing as the sentence grows in length. We conducted four experiments. The first is an ablation experiment demonstrating that the Seq2Seq + Bi-directional LSTM + Attention mechanism is superior to LSTM, Seq2Seq, Seq2Seq + Attention mechanism in natural language processing. Using an open-source Chinese corpus for testing, the accuracy was 82.1%, 63.4%, 69.2%, and 76.1%, respectively. The second experiment uses knowledge of the target domain to ask questions. Five thousand data from Taiwan Water Supply Company were used as the target training data, and a thousand questions that differed from the training data but related to water were used for testing. The accuracy of RasaNLU and this study were 86.4% and 87.1%, respectively. The third experiment uses knowledge from non-target domains to ask questions and compares answers from RasaNLU with the proposed neural network model. Five thousand questions were extracted as the training data, including chat databases from eight public sources such as Weibo, Tieba, Douban, and other well-known social networking sites in mainland China and PTT in Taiwan. Then, 1000 questions from the same corpus that differed from the training data for testing were extracted. The accuracy of this study was 83.2%, which is far better than RasaNLU. It is confirmed that the proposed model is more accurate in the general field. The last experiment compares this study with voice assistants like Xiao Ai, Google Assistant, Siri, and Samsung Bixby. Although this study cannot answer vague questions accurately, it is more accurate in the trained application fields.