Context-Matched Collage Generation for Underwater Invertebrate Detection
This addresses the challenge of training object detectors with scarce annotations in underwater imaging, though it is incremental as it builds on existing collage-based data augmentation methods.
The paper tackles the problem of limited and partially annotated training data for object detection, particularly in underwater invertebrate datasets like DUSIA, by introducing Context Matched Collages to synthesize additional training samples, resulting in state-of-the-art performance improvements with three different detectors.
The quality and size of training sets often limit the performance of many state of the art object detectors. However, in many scenarios, it can be difficult to collect images for training, not to mention the costs associated with collecting annotations suitable for training these object detectors. For these reasons, on challenging video datasets such as the Dataset for Underwater Substrate and Invertebrate Analysis (DUSIA), budgets may only allow for collecting and providing partial annotations. To aid in the challenges associated with training with limited and partial annotations, we introduce Context Matched Collages, which leverage explicit context labels to combine unused background examples with existing annotated data to synthesize additional training samples that ultimately improve object detection performance. By combining a set of our generated collage images with the original training set, we see improved performance using three different object detectors on DUSIA, ultimately achieving state of the art object detection performance on the dataset.