A new paper titled "AL-FaMoUS: Enhancing Pool-based Deep Active Learning on Imbalanced Datasets" by Zhixin Huang, Yujiang He, Marek Herde, Denis Huseljic, and Bernhard Sick was presented at the "Workshop on Interactive Adapative Learning (IAL), ECML PKDD". This research addresses the challenges of pool-based deep active learning on imbalanced datasets. AL-FaMoUS, the proposed solution, combines fast model updates and class-balanced minibatch selection to overcome degrading performance and lack of sample diversity. Experimental evaluations on various imbalanced datasets show the superiority of AL-FaMoUS over other active learning strategies and illustrate its potential to minimize the queried samples for efficient model training.

A new paper titled "AL-FaMoUS: Enhancing Pool-based Deep Active Learning on Imbalanced Datasets" by Zhixin Huang, Yujiang He, Marek Herde, Denis Huseljic, and Bernhard Sick was presented at the "Workshop on Interactive Adapative Learning (IAL), ECML PKDD". This research addresses the challenges of pool-based deep active learning on imbalanced datasets. AL-FaMoUS, the proposed solution, combines fast model updates and class-balanced minibatch selection to overcome degrading performance and lack of sample diversity. Experimental evaluations on various imbalanced datasets show the superiority of AL-FaMoUS over other active learning strategies and illustrate its potential to minimize the queried samples for efficient model training.