[1]
2022. Attention-Based Transformer Models for Real-Time Natural Language Understanding at the Edge. Journal of Intelligent Systems and Future Computing. 2, 01 (May 2022), 44–54.