In recent years, region proposals have replaced sliding windows in support of object recognition, offering more dis-criminating shape and appearance information through im-proved localization. One powerful approach for generat-ing region proposals is based on minimizing parametric en-ergy functions with parametric maxflow. In this paper, we introduce Parametric Min-Loss (PML), a novel structured learning framework for parametric energy functions. While PML is generally applicable to different domains, we use it in the context of region proposals to learn to combine a set of mid-level grouping cues to yield a small set of object region proposals with high recall. Our learning framework accounts for multiple diverse outputs, and is complemented by diversification seeds based on image location and color. This approach casts perceptual grouping and cue combina-tion in a novel structured learning framework which yields baseline improvements on VOC 2012 and COCO 2014. 1.