Woraratpanya, Kuntpong

Preferred name

Alternative Name

Weraratpanya, Kuntpong

Main Affiliation

School of Information Technology

kuntpong.wo@kmitl.ac.th

Now showing 1 - 10 of 27

Some of the metrics are blocked by your
consent settings
ADGAS: An Advanced Data Generation for Anomalous Signals
(2021-01-01)
Chalongvorachai, Thasorn
;
Woraratpanya, Kuntpong
Anomaly detection using deep learning approaches is still challenging, especially in the case of data limitations. A small number of samples for training deep learning models typically result in poor performance of detection and classification. Previously, data augmentation was one of the methods used to solve this problem. The data augmentation with rotation, permutation, time warping, and their combination can increase the performance of anomaly classification. However, this method is still limited and does not guarantee that the generated output data have adequate varieties and keep original characteristics of data. Our recent work, data augmentation and generation for anomalous time series signals (DAGAT) was proposed to expand the space of possible augmented data by implementing vanilla augmentation on various domains in conjunction with variational autoencoder (VAE). Nonetheless, the DAGAT still has barriers, which are an uncontrollable number of target results, a missed opportunity of integrating multiple augmentation characteristics in latent space, and a possibility of including any bad data for training in VAE. To overcome these limitations, this paper proposed an advanced data generation for anomalous signals (ADGAS). By focusing on the quality of generated data, one more quality classifier (QC) was added as a prepossessing step of VAE. In this way, the experimental results showed that convolutional neural networks (CNNs), used as a performance tester, trained with the generated datasets of ADGAS achieved better accuracy in classifying anomalous events when compared to models trained with a combination of rotation, permutation, and time warping data augmentation methods.
Some of the metrics are blocked by your
consent settings
Vision Transformer with Fractal Dimension Transformation: Effects of Resolution and Patch Size
(2025-01-01)
Ngamkham, Woramat
;
Woraratpanya, Kuntpong
;
Kuroki, Yoshimitsu
Vision Transformer (ViT) achieves strong performance in computer vision but requires substantial computational resources, particularly with high-resolution data. A key challenge lies in the quadratic complexity of self-attention with respect to the number of image patches, which is jointly determined by input size and patch size. Conventional resizing is a common strategy to reduce resolution and thus the number of patches, but it risks discarding structural details that may be important for prediction. To address this issue, this study investigates how input size, patch size, and dimensionality reduction influence ViT training time and prediction accuracy. Using the NIH Chest X-ray dataset, we compared two preprocessing methods: conventional resizing and a Fractal Dimension (FD)-based transformation. Results show that the FD-based method consistently reduced training time across all settings, demonstrating its effectiveness in lowering computational costs. In terms of accuracy, conventional resizing generally performed slightly better overall; however, the differences were not uniform, as smaller patches improved AUROC mainly at higher resolutions but not consistently at lower ones. These findings highlight a tradeoff between efficiency and accuracy, positioning FD-based representations as a practical complement to conventional resizing when computational resources are limited.
Some of the metrics are blocked by your
consent settings
Advancing GAN Evaluation: The Advanced Mahalanobis Distance Learning Metric for Realistic Car Damage Image Assessment
(2026-01-01)
Kyu, Phyu Mar
;
Woraratpanya, Kuntpong
Generative Adversarial Networks (GANs) have demonstrated remarkable capability in synthesizing high-quality images from limited data, addressing challenges of data scarcity and diversity in deep learning (DL) training. This is particularly valuable for car damage classification, where real-world datasets are often limited. To mitigate this, we created a custom damaged-undamaged car dataset for training GAN models and generating realistic car damage images. However, evaluating the per-image realism of GAN-generated images remains challenging. Standard GAN metrics, such as Fréchet Inception Distance (FID), Kernel Inception Distance (KID), and Inception Score (IS), provide dataset-level scores but do not assess individual image quality. Meanwhile, Image Quality Assessment (IQA) metrics require reference images, rendering them unsuitable for reference-free scenarios, particularly in unpaired GAN-generated data. To address these limitations—including the practical failure of standard Mahalanobis Distance Learning (MDL) on small or high-dimensional datasets due to non-invertible covariance matrices—we propose Advanced Mahalanobis Distance Learning (AMDL), which incorporates adaptive regularization and pseudo-inverse refinement on deep feature embeddings from pre-trained CNNs. AMDL enables stable and reliable per-image realism assessment under covariance matrix instability, without requiring large datasets or ground-truth references. Our comprehensive evaluation framework involves three procedures: (1) dataset-level evaluation of four GAN models using standard GAN metrics, (2) per-image realism assessment with AMDL, and (3) classifier-based validation with CNN and Vision Transformer (ViT) models (with vs. without AMDL). Experimental results show that AMDL provides precise per-image realism assessment, outperforms existing GAN metrics across datasets, and offers a practical solution for evaluating unpaired GAN-generated images in car damage classification.
Some of the metrics are blocked by your
consent settings
Car Damage Detection and Classification
(2020-07-01)
Kyu, Phyu Mar
;
Woraratpanya, Kuntpong
Nowadays, the proliferation of automobile industries is directly related to the increasing number of car incidents. So, insurance companies are facing many simultaneous claims and solving claims leakage. The sense of Artificial Intelligence (AI) based on machine learning and deep learning algorithms can help to solve these kinds of problem for insurance industries. In this paper, we apply deep learning-based algorithms, VGG16 and VGG19, for car damage detection and assessment in real-world datasets. The algorithms detect the damaged part of a car and assess its location and then its severity. Initially, we discover the effect of domain-specific pre-trained CNN models, which are trained on an ImageNet dataset, and followed by fine-tuning, because some of the categories can be fine-granular to get our specific tasks. Then we apply transfer learning in pre-trained VGG models and use some techniques to improve the accuracy of our system. We achieve the accuracy of 95.22% of VGG19 and 94.56% of VGG16 in the damaged detection, the accuracy of 76.48% of VGG19 and 74.39% of VGG16 in damage localization, the accuracy of 58.48% of VGG19 and 54.8% of VGG16 in damage severity with the combination of transfer learning and L2 regularization. From their results, the performance of VGG19 is better than VGG16. After analyzing and implementing our models, we find out that the results of using transfer learning and L2 regularization can work better than those of fine-tuning.
Some of the metrics are blocked by your
consent settings
Distributed compressed video sensing with a pre-learned consensus convolutional dictionary
(2026-02-27)
Muta, Ibuki
;
Weraratpanya, Kuntpong
;
Kuroki, Yoshimitsu
Distributed Compressed Video Sensing (DCVS) is a video coding framework well suited to low-power, low-complexity encoding environments. In conventional DCVS with Convolutional Sparse Representation (CSR), convolutional dictionary filters are learned from a key frame in each group of pictures (GOP), and the remaining non-key frames are reconstructed by solving a Convolutional Sparse Coding (CSC) problem using that dictionary. In this work, we investigate a CSR-based DCVS framework that instead employs a pre-learned convolutional dictionary trained offline on multiple video datasets via a consensus-based dictionary learning framework. Using this fixed dictionary, every frame in a sequence is reconstructed independently as if it were a key frame, i.e., without referencing other frames in the same sequence. We evaluate the proposed pre-learned-dictionary DCVS on the Foreman, Akiyo, and Coastguard sequences under two configurations that differ in the choice of data-fidelity term (L1 or L2) with symmetric boundary handling. Experimental results show that all test sequences can be successfully reconstructed using the pre-learned dictionary, indicating that sequence-specific key-frame-based dictionary learning at the decoder is not necessary. Moreover, the L1 data-fidelity term consistently yields better reconstruction quality than the L2 term in terms of PSNR and SSIM.
Some of the metrics are blocked by your
consent settings
Distributed compressed video sensing with multiple key frames
(2026-02-27)
Nomaguchi, Mizuki
;
Inoue, Ryota
;
Woraratpanya, Kuntpong
;
Kuroki, Yoshimitsu
Distributed Compressed Video Sensing is a video compression method utilizing Compressed Sensing and Distributed Video Coding. With this method, compressed frames are reconstructed with information obtained by applying Convolutional Sparse Coding to a non-compressed frame. In this study, we aim to increase the reconstruction accuracy by selecting multiple non-compressed frames. In addition, we use symmetric convolution in order to solve a high computational optimization problem. The experimental results show our proposed method outperforms the conventional method.
Some of the metrics are blocked by your
consent settings
Integrating Recurrent Neural Networks and Deep Q-Networks for Precision Irrigation Control
(2025-01-01)
Jogo, Fransiksus Serfian
;
Wahyunggoro, Oyas
;
Mustika, I. Wayan
;
Woraratpanya, Kuntpong
Agricultural irrigation accounts for a substantial portion of global water consumption, necessitating intelligent irrigation strategies. Conventional systems, such as fixed irrigation or evapotranspiration (ET)-based control, often exhibit inefficiencies due to reliance on static measurements and empirical farmer knowledge, lacking adaptive decision-making capabilities. To address this challenge, this paper presents an IoT-based framework integrating Recurrent Neural Networks (RNNs) for weather prediction and Deep Q-Networks (DQNs) for adaptive irrigation decisions. The RNN achieves high-precision forecasts for solar radiation, rainfall, and reference evapotranspiration (ETo) with determination coefficients (R<sup>2</sup>) of 0.819,0.915, and 0.892, respectively, surpassing the other two neural network models in our tests. Simultaneously, the DQN agent learns irrigation policies that reduce water consumption by 16.6-25.3% compared to conventional methods while maintaining post-irrigation soil moisture above the critical threshold (V<inf>mad</inf>=0.50) in testbed experiments. The proposed RNN-DQN architecture demonstrates significant improvements in water-use efficiency and plant health maintenance, offering a robust solution for smart agriculture in water-limited, data-scarce arid regions.
Some of the metrics are blocked by your
consent settings
Exploring LSTM and CNN Architectures for Sign Language Translation
(2023-01-01)
Boondamnoen, Mongkol
;
Thongsri, Kamolwich
;
Sahabantoegnsin, Thanapat
;
Woraratpanya, Kuntpong
Our study explores the application of deep learning models, specifically LSTM (Long Short-Term Memory) and CNN (Convolutional Neural Network), in the realm of sign language translation to address communication barriers faced by individuals with hearing disabilities. Using a dedicated dataset comprising ten frequently used American Sign Language words, we rigorously compare the performance of LSTM and CNN models, measuring precision and recall metrics. The LSTM model achieves a perfect accuracy score of 1, while the CNN model demonstrates a commendable accuracy of 0.9826. These results highlight the potential of these deep learning architectures to facilitate more inclusive and accessible communication avenues in sign language, bridging the communication divide.
Some of the metrics are blocked by your
consent settings
Evaluation of deep learning algorithms for semantic segmentation of car parts
(2022-10-01)
Pasupa, Kitsuchart
;
Kittiworapanya, Phongsathorn
;
Hongngern, Napasin
;
Woraratpanya, Kuntpong
Evaluation of car damages from an accident is one of the most important processes in the car insurance business. Currently, it still needs a manual examination of every basic part. It is expected that a smart device will be able to do this evaluation more efficiently in the future. In this study, we evaluated and compared five deep learning algorithms for semantic segmentation of car parts. The baseline reference algorithm was Mask R-CNN, and the other algorithms were HTC, CBNet, PANet, and GCNet. Runs of instance segmentation were conducted with those five algorithms. HTC with ResNet-50 was the best algorithm for instance segmentation on various kinds of cars such as sedans, trucks, and SUVs. It achieved a mean average precision at 55.2 on our original data set, that assigned different labels to the left and right sides and 59.1 when a single label was assigned to both sides. In addition, the models from every algorithm were tested for robustness, by running them on images of parts, in a real environment with various weather conditions, including snow, frost, fog and various lighting conditions. GCNet was the most robust; it achieved a mean performance under corruption, mPC = 35.2, and a relative degradation of performance on corrupted data, compared to clean data (rPC), of 64.4%, when left and right sides were assigned different labels, and mPC = 38.1 and rPC = 69.6 % when left- and right-side parts were considered the same part. The findings from this study may directly benefit developers of automated car damage evaluation system in their quest for the best design.
Some of the metrics are blocked by your
consent settings
3DVAE-LSTM for Extremely Rare Anomaly Signal Generation
(2022-01-01)
Kaewkiriya, Thongchai
;
Woraratpanya, Kuntpong
To overcome the uncontrolled quality output problem of data augmentation, many data generation frameworks have been proposed recently. The main concept of the data generation is for ensuring the quality of the output samples which maintain the original characteristics and highly provide the diversity of data. The benefit of this concept is improving the performance of deep learning tasks that suffer from the lack of available training samples, such as anomaly classification. Recently, 3D variational autoencoder for extremely rare case signal generation (3DVAE-ERSG) was introduced. This framework achieves the best synthesis samples for multi-class classification deep learning training. However, it is not so well applicable to sequential data. Therefore, this paper proposed a 3DVAE-LSTM framework. The new framework was replaced a VAE's feed-forward neural network with a long short-term memory (LSTM) neural network that works well with time-series signals. The experimental results show that the classification models trained with data generated by 3DVAE-LSTM have better performance than 3DVAE-ERSG in every aspect.

Woraratpanya, Kuntpong

Filters

Settings

Sort By

Results per page