How to use demo.py to get the visual feature of each region (box)?

[HaoYang0123]

I download VGG-16 model and use this model in demo.py. It returns multiple boxes in each image, however it doesn't output the visual features of these boxes. How to modify this code?

[loolzaaa]

it doesn't output the visual features of these boxes

What do you mean? Network visualize bounding boxes on image, class name and probability.

[HaoYang0123]

I have resolved this problem. I need each region feature, you can find in lib/model/faster_rcnn/faster_rcnn.py (line 86): pooled_feat = self._head_to_tail(pooled_feat). I need pooled_feat, if I use resnet-101 model, pooled_feat shape is 300*2048, which has 300 regions and each region has features (dim=2048).

[cskarthik7]

Is your torch.utils.ffi running correctly?

[KrystalCWT]

I have resolved this problem. I need each region feature, you can find in lib/model/faster_rcnn/faster_rcnn.py (line 86): pooled_feat = self._head_to_tail(pooled_feat). I need pooled_feat, if I use resnet-101 model, pooled_feat shape is 300*2048, which has 300 regions and each region has features (dim=2048).

HI, I have the same problem. Can you share your code?