Combinational sign language recognition

Liqing Gao, Wei Feng, Fan Lyu, Liang Wang

Abstract: Highlights•We propose a novel and general Combinational Sign Language Recognition (CombSLR) framework, which for the first time serves as a data augmentation method to solve the problem of the limited scale of SL data.•In CombSLR framework, we propose an EinT method, which can reliably insert E into T at the feature level to achieve the effective combination of T and E.•The extensive experiments on three public datasets demonstrate the effectiveness of our proposed EinT method, which can be embedded into any SLR model to improve its performance.