vrd-dsr
vrd-dsr copied to clipboard
Code for Visual Relationship Detection with Deep Structural Ranking (AAAI2018)
Hi, - The loss function you give in the paper differs from what the code does (it seems to me). I see that nn.MultiLabelMarginLoss is one part, but I do...
Hi, Can you share the code to extract bounding box region features for specified bounding boxes in the original image? I think this will be very useful.
I tried to run ~/tools/train.py with params --dataset vrd --loc_type 1 but got the following error. Namespace(ds_name='vrd', epochs=5, loc_type=1, lr=1e-05, model_type=None, name=None, print_freq=10, resume='', start_epoch=0, use_obj=True, use_obj_prior=True, use_so=True) Traceback (most...
Hi, I find the code to load params_emb is : Initial object embedding with word2vec with open('../data/vrd/params_emb.pkl') as f: emb_init = cPickle.load(f) net.state_dict()['emb.weight'][1::].copy_(torch.from_numpy(emb_init)) But why should use [1::]?
hello, Thanks for your code, actually, I want to try a new way to detect the visual relationship, I saw the paper there is a Fusion layer. I didn't find...
【demo.py】ImportError:cannot import name _res_detection_. I have installed "faster-rcnn.pytorch",but there is not "_res_detection_"function in "**faster-rcnn.pytorch/lib/model/utils/net_utils.py**"
I want to run this model on COCO images. I can get region proposals of those images using faster-rcnn. Apart from that, what other changes need to be in test.py?
Is it possible to change the batch size while training? I see that there is a [`_batch_size`](https://github.com/GriffinLiang/vrd-dsr/blob/5a5ae98cb4f3558c0b42df30e699e1793476dd0e/lib/data_layers/vrd_data_layer.py#L42) variable in the `VRDDataLayer` class, but the only place it is used seems...
I try to run this model on my own data, is there any data format description or script generating the data format include(.mat, .pkl etc.)