
This research aims to developed a Content-based image retrieval model for resolve semantic gaps problem where low-level features cannot correctly convey the meaning of images. The result of developed model consists of 3 modules: 1) build the image description set module, it applies a CLIP (Contrastive Language-Image Pre-training) to learn the meaning of images by self-supervised learning from the relationship between images and caption on image encoder and text encoder with cosine similarity before collecting to the image description set and create to an image feature vector. 2) query processing module to learn the meaning of the text and constructs it as a query feature vector, and 3) vectors matching module with similar values between image feature vectors and query feature vectors before sorting by relevance and display the result to the user. The result of image retrieval on the Flickr30k dataset with order-unaware metric had a mean of recall is 0.93 when the result was in the top 10 is very high, but anyway it also found that the main barrier to the accuracy of the results was image variation. When comparing the image retrieval results with the image custom dataset, it was found that the average of recall was in the same direction. And there is no problem that the model's performance is compromised when working with previously unseen data. Demonstrate that the model can retrieve content-based images effectively. It also supports users with search terms in the form of natural language that are based on the meaning of the image rather than the grammar of the language. This impact of results is a guideline for information retrieval in the future.
The research aimed (1) to develop and evaluate an emotion recognition model for Thai speech using artificial neural networks, (2) to enable accurate classification of human emotions, and (3) to bridge communication gaps between computers and users. A dataset from AIResearch.in.th consisting of 27,854 Thai-language sentences categorized into angry, sad, happy, frustrated, and neutral emotions. The Mel Frequency Cepstral Coefficients (MFCC) employed for the speech feature extraction. Data was pre-processed by augmentation techniques, including time stretching, pitch shifting, and noise injection. The pre-processed data trained for artificial neural network models, including a 1-dimensional Convolutional Neural Network (1D CNN), Long Short-Term Memory (LSTM), and a hybrid model (1D CNN and LSTM). Results showed that the hybrid model (1D CNN & LSTM) achieved the highest accuracy of 80.36%, followed by the 1D CNN model (77.52%) and the LSTM model (67.86%).
Community-based tourism focuses on enhancing the value of cultural and natural resources of the community. Tourists get an authentic experience from the local lifestyle and wisdom within the community. Therefore, having an application that recommends tourist spots personalized to individual tourists' needs, and a community tourism management system is essential. This research presents the development of an application platform for an adaptive personalized tourist attraction recommendation system (APTARS). The objective is to create an adaptive personalized tourist attraction recommendation model using a collaborative filtering method and developing cross-platform applications for tourists and entrepreneurs in community-based tourism management on Android and iOS operating systems, developed with React, Node.js, and Flutter, managing databases with MongoDB. The researcher evaluated the performance of the adaptive personalized tourist attraction recommendation model using the Mean Absolute Error value (MAE) and Root Mean Square Error (RMSE). The research findings indicate that the Euclidean distance is the most effective algorithm to measure user similarity for optimal recommendation performance. The appropriate number of neighboring nodes for travel destination recommendation is 25 nodes. The performance evaluation of the application platform by experts is rated as good. User satisfaction assessment of the application platform from a sample group of 40 people, comprising 10 community entrepreneurs and 30 tourists, found that the overall user satisfaction is rated very good (gif.latex?x\bar{}=4.52, S.D.=0.51). The accuracy and functional performance satisfaction had the highest mean (gif.latex?x\bar{}=4.56, S.D.=0.52), followed by satisfaction with the ability to work according to user needs (gif.latex?x\bar{}=4.53, S.D.=0.49), satisfaction with system usage (gif.latex?x\bar{}=4.50, S.D.=0.52), and satisfaction with system and information security (gif.latex?x\bar{}=4.48, S.D.=0.51) respectively.
The purpose of this research is to study the performance of 3D rendering software between Lumion and V-Ray plugin. It is a practical research that collects data from both brands of software to compare their performance. By comparing the performance in 3 issue and got the following results: Issue 1 Performance in 3D rendering speed it was found that Lumion program took less time to render than V-Ray plugin 14 minutes 14 seconds, representing 66.69 times. Issue 2 the complexity of use found that V-Ray plugin has less than 6 steps and Issue 3 The quality of 3D images after rendering found that in terms of noise, it was found that 3D images from V-Ray plugin noise is noticeable at 400% magnification, which is preceded by Lumion. In terms of reflection on the surface of the material with a smooth surface in noticeable at 100%, it was found that the 3D image of plugin V-Ray had obvious reflections, while the 3D image of the Lumion program did not reflect on the glass surface.
The purpose of this research were 1) to compare the critical thinking skills of students before and after engagement with an education digital game based on the phenomenon and 2) to study the fourth-grade student’s satisfactions towards education digital game. The study was conducted with a sample of fourth-grade students from two classrooms, each with 30 students, categorized by academic performances high, medium, and low. The research instruments included 1) a lesson plan for critical thinking skills based on the phenomenon, 2) the education digital game, 3) a critical thinking skills assessment, and 4) a satisfaction questionnaire. The data were analyzed by calculating means (M), standard deviations (SD), and t-test. Results indicated that 1) the average critical thinking skill scores of students who learned through education digital game were significantly higher than those who did not learn through education digital game at a significant level of .05, 2) the average critical thinking skill scores of students post-test through digital educational games were significantly higher than pre-test at a significant level of .05, and 3) The students satisfaction toward education digital game was at high level