Connecting the Semantic Dots: Zero-shot Learning with Self-Aligning Autoencoders and a New Contrastive-Loss for Negative Sampling

Autor: Mohammed Terry-Jack, Nikolai Rozanov
Rok vydání: 2022
Zdroj: 2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA).
DOI: 10.1109/icmla55696.2022.00236
Databáze: OpenAIRE