Robustness and Overfitting Behavior of Implicit Background Models
This work addresses overfitting and robustness issues in weakly supervised segmentation models for computer vision applications, presenting an incremental improvement with practical detection and mitigation strategies.
The paper investigated the overfitting behavior of image classification models modified with Implicit Background Estimation (SCrIBE), which enables weakly supervised segmentation without performance loss, and developed a label-free overfit detection criterion while evaluating the effects of data augmentations on model robustness and segmentation quality.
In this paper, we examine the overfitting behavior of image classification models modified with Implicit Background Estimation (SCrIBE), which transforms them into weakly supervised segmentation models that provide spatial domain visualizations without affecting performance. Using the segmentation masks, we derive an overfit detection criterion that does not require testing labels. In addition, we assess the change in model performance, calibration, and segmentation masks after applying data augmentations as overfitting reduction measures and testing on various types of distorted images.