Training image classifiers using Semi-Weak Label Data
This addresses a practical limitation in MIL for image classification by reducing the supervision gap, though it is incremental as it builds on existing MIL approaches.
The paper tackles the performance gap in Multiple Instance Learning (MIL) by introducing a semi-weak label learning paradigm, where class presence/absence and exact counts are known, and shows that their two-stage framework outperforms weakly supervised and learning from proportion baselines while achieving results comparable to fully supervised models on CIFAR-10.
In Multiple Instance learning (MIL), weak labels are provided at the bag level with only presence/absence information known. However, there is a considerable gap in performance in comparison to a fully supervised model, limiting the practical applicability of MIL approaches. Thus, this paper introduces a novel semi-weak label learning paradigm as a middle ground to mitigate the problem. We define semi-weak label data as data where we know the presence or absence of a given class and the exact count of each class as opposed to knowing the label proportions. We then propose a two-stage framework to address the problem of learning from semi-weak labels. It leverages the fact that counting information is non-negative and discrete. Experiments are conducted on generated samples from CIFAR-10. We compare our model with a fully-supervised setting baseline, a weakly-supervised setting baseline and learning from pro-portion (LLP) baseline. Our framework not only outperforms both baseline models for MIL-based weakly super-vised setting and learning from proportion setting, but also gives comparable results compared to the fully supervised model. Further, we conduct thorough ablation studies to analyze across datasets and variation with batch size, losses architectural changes, bag size and regularization