Vittayakorn, Sirion

Preferred name

Vittayakorn, Sirion

Main Affiliation

School of Information Technology

sirion.vi@kmitl.ac.th

Now showing 1 - 10 of 21

Some of the metrics are blocked by your
consent settings
Text and object detection on billboards
(2018-11-13)
Intasuwan, Tripidok
;
Kaewthong, Jakkraphatara
;
Vittayakorn, Sirion
To captivate people's attention in a blink of an eye, a billboard as a commercial advertisement must be attractive and informative. When space means money like on the billboard, much important information must be dropped out. Unfortunately, insufficient details and the difficulty of accessing those details might easily kill the customer's attention in a flash as well. In this work, we propose a system that automatically connects a billboard with the product website instantly. More specifically, given the billboard image as an input, the system will lead users immediately to the product website for more details, or contact information. The results demonstrate that our system is significantly outperform the Google image search baseline where the product website is always included in the top 10 websites. Finally, We posit that the system can save the customers' time and complexity to access their product of interest, resulting the increasing of product sale.
Some of the metrics are blocked by your
consent settings
Pig Carcass Assessment on Image Segmentation
(2021-01-01)
Tanthong, Jitpanu
;
Moodleah, Samart
;
Vittayakorn, Sirion
Pork is the most commonly consumed meat across the world: about one-third of all meat consumed is pork, ahead of beef and chicken. Every day a massive number of pig carcasses enter the production pipeline. When entering the pipeline, the carcasses are graded by the slaughterhouses to determine the market price of the meat. The grading criteria could depend on a variety of factors such as the tenderness, color, pH value, water holding capacity as well as the proportion of red meat inside the carcasses. Since the grade can be used to determine the market price and the commercial usage of the meat, this process is crucial. Unfortunately, the grading process is not only time consuming but also requires expertise. To mitigate this problem, in this work we propose: 1) a pig carcass image dataset segmented by experts, 2) an LSQ index image dataset and 3) an algorithm for carcass quality analysis based on the ratio of red meat inside the carcasses, or the Lenden-Speck-Quotient (LSQ). Our experimental results demonstrate that the performance of the proposed LSQ index algorithm is reliable and agrees with experts' annotation with MAPE 5.55%.
Some of the metrics are blocked by your
consent settings
Concurrent self-identification applying QR code to record class attendance (QRClass)
(2019-07-01)
Nalintipwong, Srinual
;
Tasarika, Thanarat
;
Ruksomya, Chayut
;
Vittayakorn, Sirion
;
Numnonda, Thanisa
The traditional way of recording student attendance is that an instructor calls student names one by one and records their absence or presence status. It usually consumes much time and sometimes causes errors. Therefore, some instructors may choose not to record student attendance, and this could affect academic learning and outcome. Although many self-identification methods have been explored, the entire processing time is similar to the traditional one. Hence, to solve such the problem, this research proposes a methodology to develop a concurrent class attendance system applying QR code, called QRClass. The results show that the processing time of QRClass is reduced by approximately 93.14% for 49 students compared to the traditional way. Therefore, QRClass is not only require less processing time, but also significantly increase the effectiveness of the class attendance recording method.
Some of the metrics are blocked by your
consent settings
Context-Aware Prompting for Japanese-Thai Literary Translation in a Low-Resource Setting
(2026-01-01)
Sikkhamarn, Korawit
;
Vittayakorn, Sirion
Japanese-Thai literary translation remains underexplored, particularly in low-resource settings where parallel data and evaluation resources are limited. This paper investigates whether context-aware prompting can improve Japanese-Thai literary translation using large language models. To support this study, we construct JTLitCorpus-80, a Japanese-Thai literary parallel corpus derived from 80 publicly accessible online translated web novels. Using this corpus, we compare eight prompt-based translation settings, including machine translation, post-editing, and context-aware prompting variants with and without dictionary guidance. Results on a 20-title test split show that direct context-aware prompting consistently outperforms machine translation and non-context-aware baselines, while post-editing provides only limited gains once the initial draft is already generated by the same model. We further show that prompt-derived supervision can improve a smaller open model through parameter-efficient adaptation. MQM-based human evaluation supports the automatic-metric trends, indicating that the strongest context-aware setting provides the best overall translation quality. These findings suggest that context-aware prompting is a practical strategy for Japanese-Thai literary translation in low-resource conditions.
Some of the metrics are blocked by your
consent settings
SEPBO: Trash separator bot VR game
(2021-06-30)
Theethum, Thirada
;
Iamcharoen, Sopoat
;
Arpornrat, Attawut
;
Vittayakorn, Sirion
Everyday hundreds of thousands of tons of waste goes to landfills, although more than half of it could be either recycled or composted. The waste management situation tends to get worse when there are many poorly-designed landfill sites that could either leak hazardous chemicals which can contaminate groundwater or emit harmful gases into the atmosphere. These health hazards or environmental chain problems could be diminished if everyone sorted their waste precisely. To mitigate this problems, we aim to tackle it from the beginning. Thus, in this work, we propose a computer-based VR game called SEPBO. SEPBO is an educational game which aims not only for enjoyment, but also to improve the players' waste sorting skills. The experimental results confirm that SEPBO is a better tool to improve the players' waste sorting skill compared to the interactive web-based baseline, ReCollect: The Waste Sorting Game, in ALL aspects: with 6.67% higher performance in improving players' waste sorting skill and a 48% greater degree of satisfaction than the baseline in terms of enjoyment.
Some of the metrics are blocked by your
consent settings
Facial expression recognition using local Gabor filters and PCA plus LDA
(2017-07-01)
Pumlumchiak, Tanapol
;
Vittayakorn, Sirion
A simple smile can indicate our approval, happiness or positive thoughts, while a scowl might signal displeasure or anger. Understanding facial expressions and their meanings are crucial not only in our daily life communication, but also in many applications. For example, in marketing, the customers' facial expressions indicate their response towards a product. In artificial intelligence (AI), robots can use human facial expressions as a cue for understanding their emotion in order to respond appropriately. This paper proposes a method for recognizing human facial expressions from images using local Gabor filter, Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA). The system starts by applying the face detection algorithm to detect the face from an image. From the face, the system extracts the Gabor filter responses and maps these responses into the novel feature subspace using the joined framework of PCA and LDA. Note that, in our framework, the principle component removal is also integrated into the framework. Based on the weighted neighbor approach, the system finally classifies human expressions into 4 different classes: anger, surprise, happiness and neutral. The results demonstrate that our approach significantly outperforms the baselines.
Some of the metrics are blocked by your
consent settings
AR Development for Room Design
(2018-09-06)
Reuksupasompon, Peeranut
;
Aruncharathorn, Maytichai
;
Vittayakorn, Sirion
Arranging your furniture correctly can be the difference between having a cramped, gloomy, impractical space and a cozy, elegant and functional room. Although everyone wants their place to be functional and stylish, room design demands time and expertise. While there are extensive systems for room and layout design, most of them require users to drag and drop the 3D furniture models into their 3D room, modify the furniture size/color/texture or change the camera viewpoint around to see their newly designed room from different perspectives by means of the mobile phones or tablets screen. These complications diminish the user satisfaction as well as restrict synchronous collaboration among multiple users. In this work, we propose a system for designing a furniture layout based on the augmented reality technology. Given the room floor plan and multiple QR markers, users are able to physically move their furniture (each QR marker corresponds to a 3D furniture model) around the limited space (inside the specific area defined by the room plan) to design their own functional and stylish room layout. Finally, We demonstrate that our system can not only alleviate the difficulties of existing room design systems, but also amplify the users satisfaction of room design application. Moreover, we posit that it will encourage co-design from simultaneous multiple users.
Some of the metrics are blocked by your
consent settings
Cyberbullying detection on Tweets
(2021-05-19)
Phanomtip, Aekachai
;
Sueb-In, Thaiyathorn
;
Vittayakorn, Sirion
Nowadays the young generation spends a median of 3 hours a day on social media sites such as Twitter, Facebook, Instagram, Snapchat or TikTok. As a consequence, these platforms have gradually become part of their daily life as well as an influence on their attitude and behaviour. Although the social media platforms were originally designed to share information/news or connect families and friends, they also spread fake news, conspiracy theories, hate speech and cyberbullying. Cyberbullying is bullying with the use of digital technologies which is repeated, aimed at scaring, angering or shaming those who are targeted. Feeling vulnerable, powerless, humiliated, isolated, depressed or suicidal are examples of the negative effects from cyberbullying. Unfortunately, study has shown that one-third of the students have experienced cyberbullying in their lifetime. To mitigate this problem, we propose 1) a novel dataset of 67K tweets collected from Twitter, 2) automatic annotation methods for a large-scale dataset and 3) a detection system that to identify these toxic behaviours. The experimental results demonstrate that our method outperforms the baseline by 4%.
Some of the metrics are blocked by your
consent settings
Incomplete Adventure: An Educational Game for the TOEIC Exam
(2023-01-01)
Tedsakorn, Surawee
;
Aksaralikitsanti, Nattapong
;
Vittayakorn, Sirion
These days, the Test of English for International Communication (TOEIC) plays an essential role in both the work environment and everyday life. In daily life, fluency in English offers tremendous benefits including exposure to new experiences, opportunities, cultures, and friends. In the work environment, it is an crucial factor in client relationships, prof-itability, team effectiveness, and employee engagement. Although TOEIC preparation sources can be seen everywhere, most of them aren't very attractive or motivating. As a result, it is very hard for test takers to fully focus on exam practice for a long period of time. Unfortunately, the lack of attractive and interesting of the traditional methods e.g, book or web-based exam, negatively effect the test takers' learning performance. To mitigate this problem, we propose a game-based application for TOEIC preparation called Incomplete Adventure. In game-play, player needs to answer a series of TOEIC questions to defeat the enemy without dying. Even though the experimental results demonstrate that Incomplete Adventure is comparable with traditional learning methods in improving English skill, the game receives a better degree of satisfaction than the web-based exam by 7.6%. Moreover, the overall performance score of our game also outperforms other TOEIC games by 2.68%. We posit that the enjoyment from Incomplete Adventure will enhance the motivation of learning among students and finally effect the students' perception of learning in the end.
Some of the metrics are blocked by your
consent settings
Thinkercise: An educational VR game for Python programming
(2021-05-19)
Theethum, Thirada
;
Arpornrat, Attawut
;
Vittayakorn, Sirion
Thinkercise was developed to be an educational VR game for reviewing knowledge and skills in Python programming. The game not only encourages players to move their physical body during the gameplay but also prompts players to use their basic Python programming knowledge and skill to complete a task in the game. In the gameplay, players are required to answer a series of programming related questions by moving their body in various ways in a limited time, such as punching a target, or avoiding an obstacle. To achieve the maximum score, players need to hit the correct answer targets at the right time as well as to avoid all obstacles in the game successfully. Experimental results demonstrate that our game helps increase the students' performance in computer programming by 40.8%. It outperformed self-learning by 30%.

Vittayakorn, Sirion

Filters

Settings

Sort By

Results per page