Training-Free Dataset Pruning for Instance Segmentation
This addresses the problem of inefficient dataset pruning for instance segmentation tasks, offering a faster and more effective solution for researchers and practitioners, though it is incremental as it builds on existing pruning concepts.
The paper tackles dataset pruning for instance segmentation by proposing a training-free method that uses shape and class information from annotations, achieving state-of-the-art results on VOC 2012, Cityscapes, and COCO datasets with an average 1349x speedup on COCO compared to baselines.
Existing dataset pruning techniques primarily focus on classification tasks, limiting their applicability to more complex and practical tasks like instance segmentation. Instance segmentation presents three key challenges: pixel-level annotations, instance area variations, and class imbalances, which significantly complicate dataset pruning efforts. Directly adapting existing classification-based pruning methods proves ineffective due to their reliance on time-consuming model training process. To address this, we propose a novel Training-Free Dataset Pruning (TFDP) method for instance segmentation. Specifically, we leverage shape and class information from image annotations to design a Shape Complexity Score (SCS), refining it into a Scale-Invariant (SI-SCS) and Class-Balanced (CB-SCS) versions to address instance area variations and class imbalances, all without requiring model training. We achieve state-of-the-art results on VOC 2012, Cityscapes, and COCO datasets, generalizing well across CNN and Transformer architectures. Remarkably, our approach accelerates the pruning process by an average of 1349$\times$ on COCO compared to the adapted baselines. Source code is available at: https://github.com/he-y/dataset-pruning-for-instance-segmentation