4. Interdisciplinary Conference on Electrics and Computer (INTCEC 2024) 11-13 June 2024, Chicago-USA XXX-X-XXXX-XXXX-X/XX/$XX.00 ©2024 IEEE Deep Learning Image Classification for Surat Mangyan Script Preservation Gerhard P. Tan Electronics Engineering Department Polytechnic University of the Philippines Manila, Philippines gptan@pup.edu.ph Aivor Padilla Electronics Engineering Department Polytechnic University of the Philippines Manila, Philippines aivorcpadilla@iskolarngbayan.pup.edu.ph Richmond Supleo Electronics Engineering Department Polytechnic University of the Philippines Manila, Philippines ricmondbryanasupleo@iskolarngbayan.pup.edu.ph Dianalyn Espiritu Electronics Engineering Department Polytechnic University of the Philippines Manila, Philippines dianalynpespiritu@iskolarngbayan.pup.edu.ph Romalyn Gomez Electronics Engineering Department Polytechnic University of the Philippines Manila, Philippines romalynhgomez@iskolarngbaya.pup.edu.ph Marife A. Rosales Electronics Engineering Department Polytechnic University of the Philippines Manila, Philippines marosales@pup.edu.ph Abstract— This paper presented an approach to Image Classification focusing on the Surat Mangyan script, an ancient writing system indigenous to the Philippines. Capitalizing on the Deep Learning techniques, mainly on the Convolutional Neural Networks (CNNs), the proposed system aims to accurately identify and classify Surat Mangyan characters. The research explores data augmentation methods to enhance model performance against variations in handwriting styles. The study investigates the impact of different architectural configurations on recognition accuracy, including the number of convolutional layers, kernel sizes, and activation functions. Experimental results demonstrate the effectiveness of the developed system in achieving a high classification accuracy, with a classification accuracy that accounts for 98.92%. The system also exhibits a high sensitivity (at a Macro-Averaged) of 98.98%, with an F1- score of 98.9% arriving at a precision of 98.98%. Keywords— Deep Learning, script, image recognition, Surat Mangyan, preservation I. INTRODUCTION The Hanunoo Mangyans, the largest among the eight Mangyan tribes, stand out as the "artisans of the Mangyans." with a population ranging between 15,000 to 17,000. They cultivate their sustenance through slash-and-burn farming and showcase remarkable craftsmanship, producing exquisite baskets, beadwork, and mats. Renowned for their ancient burial grounds, the Hanunoo Mangyans preserve their cultural heritage amid modernization. Originating from the ancient Sanskrit alphabet, the Hanunoos utilize the Surat Mangyan writing system, which incorporates a syllabary consisting of 18 characters comprising 3 vowels and 15 consonant-vowel combinations. Typically, this distinctive script is found inscribed on bamboo trunks, its intricate markings meticulously etched using a knife with a unique bolo shape [1]. The Mangyan script stands as a tangible testament to the rich cultural expressions and treasures embedded in Filipino heritage. Much like other Philippine scripts, it serves as a gateway to re-access, re-interpret, and re-frame our historical and cultural paradigms. Recognizing the value of indigenous scripts goes beyond preserving pre-colonial roots; it encourages a profound exploration of their content and forms, offering insights into our identity [2]. During a Zoom presentation on Language Preservation and Documentation of Hanunoo, Dr. Rowena Cristina L. Guevara, DOST Undersecretary for Research and Development, emphasized the critical importance of safeguarding the Mangyan Culture. Spearheaded by Dr. Rochelle Irene Lucas from De La Salle University, the research is dedicated to the preservation and documentation of the Hanunoo language. This endeavor was prompted by a directive issued in 2016 by Bro. Armin Luistro, the Secretary of the Department of Education (DepEd), advocating for the adoption of indigenous languages to safeguard those that are endangered or in decline. Despite its rich cultural heritage, the Hanunoo Mangyan language faces a threat to its survival due to the diminishing usage of its writing system [3]. The House of Representatives has approved a bill to protect and preserve Philippine indigenous and traditional writing systems like Baybayin. House Bill No. 10657, known as the "Philippine Indigenous and Traditional Writing Systems Act,". It mandates the inclusion of these writing systems in basic and higher education curricula and encourages awareness activities, including during Buwan ng Wika. The bill also promotes the organization of events