THAI TEXT-TO-IMAGE PROMPT ENGINEERING BY PRE-TRAINED LARGE LANGUAGE WITH STABLE DIFFUSION MODEL
DOI:
https://doi.org/10.32010/26166127.2023.6.2.171.190Keywords:
Text-to-Image Translation, Thai Prompt Engineering, Stable Diffusion Model, Image GenerationAbstract
Text-to-image (T2I) generation is a new area of large language models (LLMs), a type of prompt engineering involving inputting a textual description to generate an image. To shift a new paradigm of Thai natural language processing (Thai-NLP), this paper first presents state-of-the-art Thai Text-to-Image prompt engineering (TH-T2I) to translate Thai text into a semantic image according to the semantic Thai textual description. The pre-trained SCB-MT-EN-TH model is employed for Text-to-Text (T2T) translation. Moreover, the image generation is done according to a semantic text prompt by a stable diffusion model. The T2T is evaluated by Bi-lingual Evaluation Understudy (BLEU), while T2I is done by Inception and Frechet Inception Distance (FID). The images generated by TH-T2I were of high quality, as measured by Inception and FID. TH-T2I contributes to a T2I baseline model in Thai, preserving the Thai cultural language on digital heritage.
Downloads
References
Arreerard, R., Mander, S. & Piao, S. (2022). Survey on Thai NLP language resources and tools. In Proceedings of the 13th Conference on Language Resources and Evaluation (6495-6505). ACL.
Boonkwan, P. & Supnithi, T. (2017). Bidirectional deep learning of context representation for joint word segmentation and POS tagging. In Proceedings of the 5th International Conference on Computer Science, Applied Mathematics and Applications (184-196). Berlin, Germany: Springer.
Emsawas, T. & Kijsirikul, B. (2016). Thai Printed Character Recognition using Long Short-Term Memory and Vertical Component Shifting. In Proceedings of 14th Pacific Rim International Conference on Artificial Intelligence (106-115). Phuket, Thailand: Sprinker.
Haruechaiyasak, C., Kongthon, A., Palingoon, P., & Trakultaweekoon, K. (2013). S-Sense: A sentiment analysis framework for social media sensing. In Proceedings of the 6th International Joint Conference on Natural Language Processing (6-13). Nagoya, Japan: The Association for Computational Linguistic
Ho, J., Jain, A., & Abbeel, P. (2020). Denoising diffusion probabilistic models, arXiv: 2006.11239.
Inthajakra, L., Prachyapruit, A. & Chantavanich, S. (2016). The Emergence of communication intellectual history in Sukhothai and Ayutthaya kingdom of Thailand. Social Science Asia, 2(4), 32-41.
Ketui, N., Theeramunkong, T. & Onsuwan, C. (2013). Thai news text summarization and its application. In Proceedings of the 2013 International Symposium on Natural Language Processing, Phuket, Thailand : AIAT.
Klahan, A., Pannoi, S., Uewichitrapochana, P. & Wiangsripanawan, R. (2018). Thai word safe segmentation with bounding extension for data indexing in search engine. In Proceedings of the 14th International Conference on Computing and Information Technology (83-92). Chiang Mai, Thailand: Springer.
Koanantakool, T., Karoonboonyanan, T. & Wutiwiwatchai, C. (2009). Computers and the Thai Language. IEEE Annals of the History of Computing, 31(1), 46-61.
Kobchaisawat, T., Chalidabhongse, T. H. & Satoh, S. (2020). Scene text detection with polygon offsetting and border augmentation. Electronics, 9(1), 117.
Lapjaturapit, T., Viriyayudhakom, K. & Theeramunkong, T. (2018). Multi-Candidate word segmentation using bi-directional LSTM neural networks. In Proceedings of the 2018 International Conference on Embedded Systems and Intelligent Technology & International Conference on Information and Communication Technology for Embedded Systems (1-6). Khon Kaen, Thailand: IEEE
Lee, S., Hoover, B., Strobelt, H., Wang, Z. J., Peng, S. Y., Wright, A., Li, K., Park, H., Yang, H. & Chau, D. H. (2023). Diffusion explainer: visual explanation for text-to-image stable diffusion, arXiv: 2305.03509.
Lowphansirikul, L., Polpanumas, C., J Rutherford, A. T. & Nutanong, S. (2020). scb-mt-en-th-2020: A Large English-Thai Parallel Corpus, arXiv: 2007.03541.
Lowphansirikul, L., Polpanumas, C., J Rutherford, A. T. & Nutanong, S. (2022). A large English–Thai parallel corpus from the web and machine-generated text. Language Resources and Evaluation. 56(2), 477-499.
Lowphansirikul, L., Polpanumas, C., Jantrakulchai, N. & Nutanong, S. (2021). WangchanBERTa: Pretraining transformer-based Thai Language Models, arXiv: 2101.09635.
Mookdarsanit, L. & Mookdarsanit, P. (2019a). SiamFishNet: The deep investigation of Siamese fighting fishes. International Journal of Applied Computer Technology and Information Systems, 8(2), 40-46.
Mookdarsanit, L. & Mookdarsanit, P. (2019b). Thai herb identification with medicinal properties using convolutional neural network. Suan Sunandha Science and Technology Journal, 6(2), 34-40.
Mookdarsanit, L. & Mookdarsanit, P. (2020a). An adversarial perturbation technique against reCaptcha image attacks. Journal of Science and Technology Buriram Rajabhat University, 4(1), 33-45.
Mookdarsanit, L. & Mookdarsanit, P. (2020b). The insights in computer literacy toward HR intelligence: some associative patterns between IT subjects and job positions. Journal of Science and Technology RMUTSB, 4(2), 12-23 .
Mookdarsanit, L. & Mookdarsanit, P. (2021a). Combating the hate speech in Thai textual memes. Indonesian Journal of Electrical Engineering and Computer Science, 21(3), 1493-1502.
Mookdarsanit, L. & Mookdarsanit, P. (2021b). ThaiWritableGAN: Handwriting generation under given information. International Journal of Computing and Digital Systems, 10(1), 689-699.
Mookdarsanit, L. & Mookdarsanit, P. (2022). Thai NLP-based Text Classification of the 21st-century Skills toward Educational Curriculum and Project Design. International Journal of Applied Computer Technology and Information Systems, 11(2), 62-67.
Mookdarsanit, L. & Mookdarsanit, P. (2023). The cosmetic surgery recommendation: Facial acne localization and recognition. International Journal of Applied Computer Technology and Information Systems, 12(2), 1-6.
Mookdarsanit, L. (2020). The intelligent genuine validation beyond online Buddhist amulet market. International Journal of Applied Computer Technology and Information Systems, 9(2),7-11.
Mookdarsanit, P. & Mookdarsanit, L. (2018a). A content-based image retrieval of Muay-Thai folklores by salient region matching. International Journal of Applied Computer Technology and Information Systems, 7(2), 21-26.
Mookdarsanit, P. & Mookdarsanit, L. (2018b). An automatic image tagging of Thai dance’s gestures. In Proceedings of Joint Conference on ACTIS & NCOBA (76-80). Ayutthaya, Thailand.
Mookdarsanit, P. & Mookdarsanit, L. (2018c). Contextual image classification towards metadata annotation of Thai-tourist attractions. ITMSoc Transactions on Information Technology Management, 3(1), 32-40.
Mookdarsanit, P. & Mookdarsanit, L. (2018d). Name and recipe estimation of Thai-desserts beyond image tagging. Kasem Bundit Engineering Journal, 8(Special Issue), 193-203.
Mookdarsanit, P. & Mookdarsanit, L. (2019). TGF-GRU: A cyber-bullying autonomous detector of lexical Thai across social media. NKRAFA Journal of Science and Technology, 15, 50-58.
Mookdarsanit, P. & Mookdarsanit, L. (2020a). Thai-IC: Thai image captioning based on CNN-RNN architecture. International Journal of Applied Computer Technology and Information Systems, 10(1), 40-45.
Mookdarsanit, P. & Mookdarsanit, L. (2020b). ThaiWrittenNet: Thai handwritten script recognition using deep neural networks. Azerbaijan Journal of High Performance Computing, 3(1), 75-93.
Mookdarsanit, P. & Mookdarsanit, L. (2020c). The autonomous nutrient and calorie analytics from a Thai food image. Journal of Faculty Home Economics Technology RMUTP, 2(1), 1-12.
Mookdarsanit, P. & Mookdarsanit, L. (2021a). PhosopNet: An improved grain localization and classification by image augmentation. TELKOMNIKA Telecommunication, Computing, Electronics and Control, 19(2), 479-490.
Mookdarsanit, P. & Mookdarsanit, L. (2021b). The COVID-19 fake news detection in Thai social texts. Bulletin of Electrical Engineering and Informatics, 10(2), 988-998.
Mookdarsanit, P. & Rattanasiriwongwut, M. (2017a). GPS determination of Thai-temple arts from a aingle photo. In Proceedings of 11th International Conference on Applied Computer Technology and Information Systems (42-47). Bangkok, Thailand.
Mookdarsanit, P. & Rattanasiriwongwut, M. (2017b). Location estimation of a photo: a Geo-signature MapReduce workflow. Engineering Journal, 21(3), 295-308.
Mookdarsanit, P. & Rattanasiriwongwut, M. (2017c). MONTEAN Framework: a magnificent outstanding native-Thai and ecclesiastical art network. International Journal of Applied Computer Technology and Information Systems, 6(2), 17-22.
Rombach, R., Blattmann, A., Lorenz, D., Esser, P. & Ommer, B. (2021). High-resolution image synthesis with latent diffusion models, arXiv: 2112.10752.
Ruangrajitpakorn, T. (2006). An example-based machine translation: a case study of translating stock reports from thai to english [Master’s thesis, Chulalongkorn University]. Graduate School, Chulalongkorn University.
Soimart, L. & Mookdarsanit, P. (2016). Gender estimation of a portrait: Asian facial-significance framework. In Proceedings of the 6th International Conference on Sciences and Social Sciences. Mahasarakham, Thailand.
Soimart, L. & Mookdarsanit, P. (2017a). Ingredients estimation and recommendation of Thai-foods. SNRU Journal of Science and Technology, 9(2), 509-520.
Soimart, L. & Mookdarsanit, P. (2017b). Name with GPS auto-tagging of Thai-tourist attractions from an image. In Proceedings of the 2nd Technology Innovation Management and Engineering Science International Conference (211-217). Nakhon Pathom, Thailand.
Sornlertlamvanich, V. (2019). Natural language processing research in Thai context - A 29-year journey of Thai NLP. Retrieved from: https://www.slideshare.net/virach/nlp-historythaivirach20191025
Sriwirote, P., Thapiang, J., Timtong, V. & Rutherford, A. T. (2023). PhayaThaiBERT: Enhancing a Pretrained Thai Language Model with Unassimilated Loanwords, arXiv: 2311.12475.
Sutthaluang, N. & Prakancharoen, S. (2020). Prediction and protection of car driving accident in urban zone. International Journal of Innovation, Creativity and Change, 14(8), 308-336.
Sutthaluang, N. (2019). An open library development for pesticide residue analytics in vegetables. International Journal of Applied Computer Technology and Information Systems, 8(2), 31-36.
Taerungruang, S. & Aroonmanakun, W. (2018). Constructing an academic Thai plagiarism corpus for benchmarking plagiarism detection systems. GEMA Online Journal of Language Studies, 18(3), 186-202 .
Tapsai, C., Unger, H. & Meesad, P. (2020). The application of Thai natural language processing. Thai Natural Language Processing, 1, 131-159.
Theeramunkong, T., Sornlertlamvanich, V., Tanhermhong,T. & Chinnan, W. (2000). Character cluster based Thai information retrieval. In Proceedings of the 2000 International Workshop on Information Retrieval with Asian Languages, (75-80), Hong Kong, China : ACM.
Tirasaroj, N. (2016). A study of word sense discrimination in Thai using latent semantic analysis [Doctoral dissertation, Chulalongkorn University]. Graduate School, Chulalongkorn University.
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L. & Polosukhin, I. (2017). “Attention Is All You Need,” In Proceedings of the 2017 International Conference on Neural Information Processing Systems, (6000-6010). Long Beach, California : ACM.
