search

batra-mlp-lab / visdial

[CVPR 2017] Torch code for Visual Dialog
https://arxiv.org/abs/1611.08669
Other
227 stars 69 forks source link

Support for extracting conv layer / spatial features for ResNets (for attention-based models) #38

Open abhshkdz opened 6 years ago

abhshkdz commented 6 years ago

Needs a command-line argument (like the VGG models) and corresponding model:remove() calls here: https://github.com/batra-mlp-lab/visdial/blob/master/data/prepro_img_resnet.lua#L37