search

fluque1995 / tfm-anomaly-detection

Master's thesis in Data Sciences. Comparison and proposal of enhancement for "Real-world anomaly detection in surveillance videos"
GNU General Public License v3.0
24 stars 7 forks source link

Resources cosuming #2

Open SaadTazroute opened 3 years ago

SaadTazroute commented 3 years ago

Hello,

Thanks for the repo, I have some remarks to make to improve your rendering, and to be able to run it on less energy consuming machines.

For the Original code, the features extraction ("extract_features.py") is too much consuming resources (Ram crash)

in the proposal folder, the weight files of the feature extractor are seemingly badly saved.

The Videogenerator class that you could write based on the code of \url{https://github.com/metal3d/keras-video-generators/tree/master/src/keras_video} is very simplistic, other parameters can be included in the class namely (use_frame cache) in order to optimize the memory management.

atchelodev commented 7 months ago

Hello i have the same problem , can you please , i'm trying to run it on google colab but i don't have enough ressources , can you please explain me clearly how to improve . thanks