Researcher Nguyen Khanh Duy Participates in and Presents at MAPR 2026

The conference brought together a large number of researchers, lecturers, specialists, and students from Vietnam and abroad to discuss and share perspectives and new research findings in Artificial Intelligence, Machine Learning, Pattern Recognition, Computer Vision, Multimedia Analysis, and related fields. Notably, the conference was also attended by Professor Min-Te Sun, researcher Nguyễn Khánh Duy’s academic advisor, from National Central University, Taiwan. The participation of professors and experts from various countries demonstrated the professionalism and reputation of MAPR 2026 in the field of computer science.

Researcher Khánh Duy presented a paper entitled “Scene Text Detection Based on Attention ConvNeXt.” The paper focuses on applying Artificial Intelligence (AI) to scene text detection, such as text appearing on signboards, traffic signs, and curved text. Although existing architectures, including CNN-based and Transformer-based models, have achieved promising performance in this field, they still require substantial computational resources. To address this issue, the authors developed a new model based on the FAST framework. By introducing the Attention Next Block and integrating Weight Standardization (WS), Batch-Channel Normalization (BCN), and the SiLU activation function into the architecture, the proposed model achieved strong performance on the MSRA-TD500 (90.1 F1-score), Total-Text (90.5 F1-score), and SCUT-CTW1500 (87.5 F1-score) benchmarks with only 30.27 million parameters, compared with 36.59 million parameters for MixNet and 47.44 million for DPText-DETR. The experiments demonstrated that the proposed model offers significant efficiency advantages while requiring fewer computational resources.

In addition to oral presentations, MAPR 2026 featured a poster exhibition showcasing conference papers. The exhibition took place in an engaging atmosphere, with students and researchers actively participating in question-and-answer sessions. These discussions provided substantial benefits to researchers while also helping inspire students’ interest and passion for scientific research.

The event provided an excellent opportunity to connect with experts from around the world, helping strengthen confidence in and enhance the presence of the Artificial Intelligence Center of Tra Vinh University at professional and academic forums. It also established a strong foundation for seeking potential research collaborations and developing new research ideas in the future.