QazLip: A Dataset of Lip Movements for Command Recognition
收藏NIAID Data Ecosystem2026-05-10 收录
数据链接：
官方服务：
资源简介：
The dataset contains 102 frequently used Kazakh nouns with visually distinct articulation, spoken by 26 participants of different ages, resulting in approximately 34,000 short video samples. Recordings were processed at 60 FPS, resulting in around 1,200,000 frames. Each word was repeated 100-400 times by various speakers to ensure articulation variability.
创建时间：
2025-09-17
