FenceMask: A Data Augmentation Approach for Pre-extracted Image Features
This addresses data augmentation challenges in computer vision, particularly for small objects and fine-grained tasks, but appears incremental as it builds on existing occlusion simulation strategies.
The paper tackles the problem of data augmentation for pre-extracted image features by proposing FenceMask, a method based on simulating object occlusion to balance occlusion and information retention, which achieved significant performance improvements on fine-grained visual categorization tasks and the VisDrone dataset.
We propose a novel data augmentation method named 'FenceMask' that exhibits outstanding performance in various computer vision tasks. It is based on the 'simulation of object occlusion' strategy, which aim to achieve the balance between object occlusion and information retention of the input data. By enhancing the sparsity and regularity of the occlusion block, our augmentation method overcome the difficulty of small object augmentation and notably improve performance over baselines. Sufficient experiments prove the performance of our method is better than other simulate object occlusion approaches. We tested it on CIFAR10, CIFAR100 and ImageNet datasets for Coarse-grained classification, COCO2017 and VisDrone datasets for detection, Oxford Flowers, Cornel Leaf and Stanford Dogs datasets for Fine-Grained Visual Categorization. Our method achieved significant performance improvement on Fine-Grained Visual Categorization task and VisDrone dataset.