作者: Kripabandhu Ghosh , Saptarshi Ghosh , Anurag Roy , Vinay Kumar Verma
DOI:
关键词: Training set 、 Computer science 、 Pattern recognition 、 Image retrieval 、 Shot (filmmaking) 、 Representation (mathematics) 、 Expectation–maximization algorithm 、 Benchmark (computing) 、 Hash function 、 Artificial intelligence 、 Zero (linguistics)
摘要: Most existing algorithms for cross-modal Information Retrieval are based on a supervised train-test setup, where a model learns to align the mode of the query (eg, text) to the mode of …