Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/anurima-saha/multilabel_image_classification


https://github.com/anurima-saha/multilabel_image_classification

classification cnn-pytorch deep-learning deep-neural-networks image-classification multi-label-classification

Last synced: about 18 hours ago
JSON representation

Awesome Lists containing this project

README 

        
# Multilabel_Image_Classification

In traditional classification tasks, each instance is associated with a single label, representing a mutually exclusive relationship between classes. However, in many real-world applications an instance can belong to multiple classes simultaneously.This requires predicting multiple, potentially correlated labels for each instance. In this project, we have utilized the power of computer vision and deep learning  to experiment with multi-label image classification tasks. Moving away from the traditional idea of image classification, we drew inspiration from real-world scenarios where images are typically a mix of diverse visual elements. We have navigated through this complex task leveraging  different CNN based architectures and transfer learning. We went further to use the GCN superimposition on CNN and track the improvement in model efficiency.

As a part of our work, we have experimented on movie posters data trying to make multi-label predictions of respective genres. The goal is to compare the performance of two different CNN algorithms(VGG16 and Resnet50) across the same dataset with a standardized amount of training. We have also trained a superimposition of GCN on CNN using  the same data and traced the model performance.