Alberto Montes

I cannot reproduce myself (very extensively). I don't think add a dependency is the solution. I also think this is not critical at all as the public release is right...

I have tested to do it with pip 10.0.1 too. I think that you might have updated pip before updated davisinteractive in the conda environment and this can be the...

I know that the paper doesn't mention any preprocessing, but the typical preprocessing used in deep learning with images and video is substracting the mean of each of the color...

To solve this it requires to add the src folder into `PYTHONPATH`: `export PYTHONPATH='$PYTHONPATH:/path/to/src'`

This means that OpenCV is not able to open the file. The usual setup is to create a virtual environment with `--system-site-packages` so if you have OpenCV installed on your...

Which version of Numpy are you using? I have checked and with version 1.11.2 should work. If not try to print the values of `s`, `e` and `activity_idx` to console...

I experienced the same while running this, and I wasn't able to find the cause (as there is concurrency to fetch the videos, maybe at some point or some video...

Hi, I can not recall the exact versions of the CUDA/cudnn software used to run this. I only remember I used a Titan X to do the training. Take into...

To train your own datasets it would be necessary to dig into the code and overwrite the functions which read the videos in the first place to read the videos...

I think this is a design decision that might be needed to be made (I am not leaning to any specific option here): 1. We define the attributes, their options...