Reduced Focal Loss: 1st Place Solution to xView object detection in Satellite Imagery
This addresses the challenge of object detection in satellite imagery with imbalanced data, though it is incremental as it builds on existing focal loss methods.
The paper tackled the problem of highly imbalanced object classes in the DIUx xView 2018 satellite imagery dataset by introducing a novel Reduced Focal Loss function, which achieved 1st place in the challenge.
This paper describes our approach to the DIUx xView 2018 Detection Challenge [1]. This challenge focuses on a new satellite imagery dataset. The dataset contains 60 object classes that are highly imbalanced. Due to the imbalanced nature of the dataset, the training process becomes significantly more challenging. To address this problem, we introduce a novel Reduced Focal Loss function, which brought us 1st place in the DIUx xView 2018 Detection Challenge.