Li Yao
You guys are doing something really fancy, way beyond my league to grasp. On Thu, Apr 30, 2015 at 3:54 PM, Bart van Merriënboer < [email protected]> wrote: > @yaoli https://github.com/yaoli...
The splits are already shuffled.
"for each video are less than actual total no. of frames in the video", I think they are the same, unless you change the sampling rate during frame extraction with...
Basically one needs to follow the preprocessing steps mentioned in the paper, turning a new video clip into a series of frames, each represented by a 1024 dim vector.
The code is only with 2D CNN features, so no need to extract 3D CNN features.
I didn't install git-lfs, not sure what it's actually doing. Why does loading pkl fail without it? I'm using a Linux machine.
Oh, I see. That's why. Thanks a lot. I'll try it later.
It works. Thanks a lot. From what I can see in those pkl files, videos on TACoS are further chopped into small short clips. To be consistent, it would be...
Thanks very much indeed.
> {'questions': [{'T': 'person put pan on _ oven .', 'W3': 'leek', 'W2': 'start', 'W1': 'unwrap', 'W0': 'heat'}, {'T': 'person put _ on heat oven .', 'W3': 'proceed', 'W2': 'pull...