0

I'm trying to get this repo to work. I followed the instruction and get the sample data using this script (taken from the same repo):

#!/usr/bin/env sh
# This script downloads the trained S2VT VGG (RGB) model,
# associated vocabulary, and frame features for the validation set.

echo "Downloading Model and Data [~400MB] ..."

wget --no-check-certificate https://www.dropbox.com/s/wn6k2oqurxzt6e2/s2s_vgg_pstream_allvocab_fac2_iter_16000.caffemodel
wget --no-check-certificate https://www.dropbox.com/s/20mxirwrqy1av01/yt_allframes_vgg_fc7_val.txt
wget --no-check-certificate https://www.dropbox.com/s/v1lrc6leknzgn3x/yt_coco_mvad_mpiimd_vocabulary.txt

echo "Organizing..."

DIR="./snapshots"
if [ ! -d "$DIR" ]; then
    mkdir $DIR
fi
mv s2s_vgg_pstream_allvocab_fac2_iter_16000.caffemodel $DIR"/s2vt_vgg_rgb.caffemodel"

echo "Done."

In the next step they said that I need to sample video frames and extract VGG features for the frames. I'm not exactly sure how to do this. I have followed the instruction on Caffe but the features are not in the same format.

So how do I extract the VGG features in the same format as the yt_allframes_vgg_fc7_val.txt?

stop-cran
  • 4,229
  • 2
  • 30
  • 47
Diep TL
  • 163
  • 1
  • 1
  • 5
  • 1
    Possible duplicate of [Extracting Features from VGG](http://stackoverflow.com/questions/33208282/extracting-features-from-vgg) – Prune Jul 08 '16 at 18:00

1 Answers1

0

This repo provides the script for extracting VGG features from videos - https://github.com/jesu9/VGGFeatExtract

In particular, see the script video_demo.py. This will output mat files which you will have to convert to txt files.

VGG 16 layer model and prototxt files available at - https://gist.github.com/ksimonyan/211839e770f7b538e2d8

vinsin
  • 1
  • 3