Skip to content

Multi-Class and Multi-Label Classification on an unbalanced Image dataset.

License

Notifications You must be signed in to change notification settings

t0re199/AIEV_PROJECT

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 

Repository files navigation

README

This project has been developed for the Images and Videos Analysis class by t0re199 and adel3roman6.

The project required to perform a Multi-Class and Multi-Label Classification on an unbalanced dataset. In particulat, the dataset contains frames of film trailer (the number of frames can be different for each film).

In this project the well-known image classification architectures ResNet34 and Vgg-16 have been used. Since PyTorch has been used as Gradient Computing Library, the above cited architecture were taken from the TorchVision Library.

Since the dataset contains unlabeled data too, in order to improve the model's extracted features quality a self-supervised learning strategy has been applied during the last epochs of the training process.

Releases

No releases published

Packages

No packages published