Unsupervised Learning Approaches for Zooplankton Classification: Recent Trends and Advances

Zooplankton are key components of the aquatic food web and present a lot of taxonomic diversity. Over the years, various Machine Learning techniques have been employed for the classification of zooplankton. Supervised Learning has been widely utilised in zooplankton classification, presenting commendable performance. However, it requires a substantial amount of manually labelled images, and the volume of collected zooplankton images is extensive. Consequently, Unsupervised Learning techniques have proven exceptionally valuable and popular for clustering unlabelled zooplankton data. Our study compiles and elucidates the Unsupervised Learning methods applied to zooplankton classification, while also comparing their performance and popularity over the years, based on observations from respective experimental studies.