Abstract: Few-shot object detection achieves rapid detection of novel-class objects by training detectors with a minimal number of novel-class annotated instances. Transfer learning-based few-shot ...
Abstract: Conventional training of deep neural networks requires a large number of the annotated image which is a laborious and time-consuming task, particularly for rare objects. Few-shot object ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback