search

LisaAnne / LocalizingMoments

Github for my ICCV 2017 paper: "Localizing Moments in Video with Natural Language"
188 stars 44 forks source link

Question about extract feature #12

Closed yuxinyan closed 6 years ago

yuxinyan commented 6 years ago

Hi, Lisa. Could you tell me what model you use to extract RGB and Flow feature from each frame? VGG16 or VGG19 pretrained on ImageNet?