A Coreset Selection Framework with Ensemble Aggregation for Image Classification
This work addresses the problem of efficient training data selection for image classification, but the improvements are incremental over existing methods.
The paper proposes a coreset selection framework with ensemble aggregation to reduce training data size for image classification. SCOSS, a score-stratified sampling method, combined with ensemble predictions, achieves competitive accuracy with baselines and favorable accuracy-efficiency trade-offs, e.g., SGC with SCOSS outperforms SVMs on a fine-grained dataset with fewer labeled samples.
The rapid growth of image data has produced large-scale datasets, raising concerns about the time and memory costs of model training. Selecting representative training subsets, however, remains challenging: individual sample contributions are unclear, and model behavior varies across datasets and runs. We address these challenges with a framework that combines coreset selection with an ensemble aggregation over multiple runs. For coreset selection, we propose SCOre-Stratified Selection (SCOSS), which partitions the training data into intervals based on a chosen score and samples from each interval. The ensemble combines predictions from multiple runs, each performed on an independently sampled training subset. As baselines, we use moderate and random selection, each in original and class-balanced versions. We assess the framework with Simple Graph Convolution (SGC) and Support Vector Machine (SVM) classifiers under different sampling ratios. Experiments show that SCOSS is competitive with baselines, often the best choice for SGC, and enables favorable trade-offs between accuracy and efficiency. On the fine-grained dataset, SGC with SCOSS outperforms SVMs when using fewer labeled samples. The code and supplementary materials are publicly available at http://scoss.lucasvalem.com.