Nikolay

Hi @chengshuai My testing stats pascal_voc 2007(trian:2007+2012,test:2007): **pvanet_obsolete(pva9.0)** 100k MeanAP=0.7190 **PVA 9.1** 100k Mean AP = 0.7512 120k Mean AP = 0.7768 360k Mean AP = 0.7922 **ResNet50** 80k Mean...

@chengshuai Main idea that you use wrong weights caffemodel (for pva 9.0). Need to load pretrained from imagenet with batchnorms and not compressed. Smaller resault on pva 9.1 may have...

@samuel1208 It is standard faster-rcnn pipeline, and if you run from python it will reshape model to cfg.TEST.SCALES = (640,) size. For my tests it runs with different sizes.

Thank you. Visualization with `tr_boxes = bbox_transform_inv(boxes, 0.1*bbox_targets)` is correct. I miss this point. But not understand why they not change loss_weight at 'bbox loss'.

Hi I set minsize=30 %minimum face size And it runs 20fps-80fps

@zHanami sorry for disinformation I run test on CelebA with min_face=60 and it is 20fps-80fps Test on wider with min_face=30 is 5.5fps as your resault Cascade like MTCNN potential more...