Bounding-Box Labeling, Training and Detection in MIPAR?

Bounding-Box Labeling, Training and Detection in MIPAR?

Recently, there's conversation at workplace to explore bounding-box detection of objects in photo-images. I understand that MIPAR's current DL technology is for pixel-level semantic segmentation. I wonder if MIPAR has plans to develop bounding-box labeling, training and detection capabilities? Thanks.