KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/kmitl/1

Browse

Settings

Author: search.filters.author.Marukatat, Sanparith

Search Results

Now showing 1 - 5 of 5
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Quantitative Ultrasound Assessment of Liver Fat Using Deep Learning and Clinical Data Integration
    (2025-12-01)
    Jamrasnarodom, Jirakorn
    ;
    Apiparakoon, Terapap
    ;
    Marukatat, Sanparith
    ;
    Chaichuen, Oracha
    ;
    Sukchareon, Sasima
    Purpose: The increasing prevalence of Metabolic Dysfunction-Associated Steatosis Liver Disease (MASLD) highlights the need for effective assessment tools, particularly for the quantitative evaluation of hepatic steatosis. Given the limited availability of transient elastography (TE), especially in resource-limited settings, we aimed to develop an Artificial Intelligence (AI) model to quantify hepatic steatosis using conventional ultrasound, which is widely available in most healthcare facilities. Methods: Liver ultrasonographic images and Controlled Attenuation Parameter (CAP) scores obtained from TE were collected from patients between 2017 and 2023. A predictive model was developed by integrating YOLOv8 for image classification with Principal Component Analysis and Lasso regression to estimate CAP scores from the ultrasonographic images. The dataset was randomly divided into training (80%), validation (10%), and test (10%) sets. Baseline patient characteristics and laboratory data were also incorporated to enhance model performance. The model’s predictive ability was evaluated using the coefficient of determination (R²) and mean squared error (MSE). Results: A total of 1065 images from 352 patients were included. The initial model achieved an R² of 0.55 and an MSE of 1004.07. Subgroup analysis revealed that the right intercostal view yielded the best performance (R²=0.74, MSE = 637.99). After incorporating patient characteristics and laboratory data, the model’s performance improved significantly (R²=0.90, MSE = 245.79). Conclusion: The AI-assisted model showed promise for accessible and non-invasive assessment of hepatic steatosis, particularly when using the right intercostal view and supplemental clinical data. Further validation is warranted to improve its accuracy and generalizability.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    ThaiTC:Thai Transformer-based Image Captioning
    (2022-01-01)
    Jaknamon, Teetouch
    ;
    Marukatat, Sanparith
    For problems with image captioning is a technique that has been used for a long time. In the past, there was a way to use convolutional neural network (CNN) for feature extraction and recurrent neural network (RNN) for generating text, and especially in Thai language, It has to be developed further in the era of the popular use of transformers. This paper proposes an end-to-end image captioning with pretrained vision Transformers (ViT) and text transformers in Thai language models namely ThaiTC, Which leverages the transformer architecture both. We has experiment pretrained vision transformer and text transformer in Thai language that best for Thai image captioning and tested on 3 Thai image captioning datasets 1) Travel 2) Food 3) Flickr 30k(tr anslate) with different challenges. Includes freeze vision transformers weight training for image captioning dataset training with less image features, From the experiment, We found that ThaiTC performed much better in the Food and Flickr30k datasets than the Travel datasets, Which allowed us to automatically create subtitles about food and travel.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Robust speech recognition using noise-cluster HMM interpolation
    (2008-12-01)
    Thatphithakkul, Nattanun
    ;
    Kruatrachue, Boontee
    ;
    Wutiwiwatchai, Chai
    ;
    Marukatat, Sanparith
    ;
    Boonpiam, Vataya
    This paper proposes a novel approach called noisecluster HMM interpolation for robust speech recognition. The approach helps alleviating the problem of speech recognition under noisy environments not trained in the system. In this method, a new HMM is interpolated from existing noisy-speech HMMs that are best matched to the input speech. This process is performed on-the-fly with an acceptable delay time and, hence, no need to prepare and store the final model in advance. Interpolation weights among HMMs can be determined by either a direct or a tree-structured search. Evaluated focusing on speech in unseen noisy-environments, the proposed method obviously outperforms a baseline system whose acoustic model for such unseen environment is selected from a tree structure. © 2008 IEEE.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Tree-structured model selection and simulated-data adaptation for environmental and speaker robust speech recognition
    (2007-12-01)
    Thatphithakkul, Nattanun
    ;
    Kruatrachue, Boontee
    ;
    Wutiwiwatchai, Chai
    ;
    Marukatat, Sanparith
    ;
    Boonpiam, Vataya
    This paper proposes the use of tree-structured model selection and simulated-data in maximum likelihood linear regression (MLLR) adaptation for environment and speaker robust speech recognition. The objective of this work is to solve major problems in robust speech recognition system, namely unknown speaker and unknown environmental noise. The proposed solution is composed of two components. The first one is based on a tree-structured model for selecting a speaker-dependent model that best matches to the input speech. The second component uses simulated-data to adapt the selected acoustic model to fit with the unknown noise. The proposed technique can thus alleviate both problems simultaneously. Experimental results show that the proposed system achieves a higher recognition rate than the system using only the input speech in adaptation and the system using a multi-conditioned acoustic model. © 2007 IEEE.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    A simulated-data adaptation technique for robust speech recognition
    (2006-01-01)
    Thatphithakkul, Nattanun
    ;
    Kruatrachue, Boontee
    ;
    Wutiwiwatchai, Chai
    ;
    Marukatat, Sanparith
    ;
    Boonpiam, Vataya
    This paper proposes an efficient acoustic model adaptation method based on the use of simulated-data in maximum likelihood linear regression (MLLR) adaptation for robust speech recognition. Online MLLR adaptation is an unsupervised process which requires an input speech with phone labels transcribed automatically. Instead of using only the input signal in adaptation, our proposed simulated data method increases the size of adaptation data by adding noise portions extracted from the input speech to a set of pre-recorded clean speech, whose correct transcriptions are known. Various configurations of the proposed method are explored. Evaluations are performed with both additive and real noisy speech. The experimental results show that the proposed system achieves higher recognition rate than the system using only the input speech in adaptation and the system using a multi-conditioned acoustic model.