Transactional Filipino Sign Language Dataset

This dataset contains 1,065 .npy coordinates from video samples of Filipino Sign Language (FSL) gestures specifically curated for transactional and commercial interactions. The data was collected to support the development of Sign Language Recognition (SLR) systems aimed at bridging communication gaps in public facilities, markets, and government offices. The dataset consists of 26 distinct classes including numbers (1–10) and common transactional phrases: Yes, No, Thank You, Welcome, Again, How Much, How Many, Coin, Cash, Please, Wait, Card, Receipt, Discount, Hello, and Problem. This dataset is suitable for training and validating deep learning models (such as LSTM, GRU, or Transformers) for dynamic gesture recognition, particularly for edge computing applications in accessibility technology.本数据集包含1065份菲律宾手语（Filipino Sign Language, FSL）手势视频样本的.npy格式坐标数据，专为交易与商务交互场景定制筛选。本数据集的采集旨在支持手语识别（Sign Language Recognition, SLR）系统的研发，该系统旨在弥合公共设施、市场与政府办公场所中的沟通鸿沟。本数据集涵盖26个独立类别，包含数字1至10以及常用交易类短语：是、否、谢谢、欢迎、再次、多少钱、多少、硬币、现金、请、稍候、卡片、收据、折扣、你好以及问题。本数据集适用于训练与验证面向动态手势识别的深度学习模型（如长短期记忆网络LSTM、门控循环单元GRU及Transformer），尤其适用于无障碍技术中的边缘计算应用场景。