Realistic Model Selection for Weakly Supervised Object Localization

arXiv:2404.10034v1 Announce Type: cross
Abstract: Weakly Supervised Object Localization (WSOL) allows for training deep learning models for classification and localization, using only global class-level labels. The lack of bounding box (bbox) supervision during training represents a considerable challenge for hyper-parameter search and model selection. Earlier WSOL works implicitly observed localization performance over a test set which leads to biased performance evaluation. More recently, a better WSOL protocol has been proposed, where a validation set with bbox annotations is held out …

abstract arxiv box challenge class classification cs.cv cs.lg deep learning global labels localization model selection object performance search supervision training type