Javascript must be enabled for the correct page display

Using Intersection over Union loss to improve Binary Image Segmentation

van Beers, Floris (2018) Using Intersection over Union loss to improve Binary Image Segmentation. Bachelor's Thesis, Artificial Intelligence.


Download (256kB) | Preview
[img] Text
Restricted to Registered users only

Download (94kB)


In semantic segmentation tasks the Jaccard Index, or Intersection over Union (IoU), is often used as a measure of success. While this measure is more representative than per-pixel accuracy, state-of-the-art neural networks are still trained on accuracy by using Binary Cross Entropy Loss. In this research, an alternative is used where a neural network will be trained for a segmentation task on face detection by optimizing directly on an approximation of IoU. When using this approximation, IoU becomes differentiable and can be used as a loss function. The comparison between IoU loss and Binary Cross Entropy loss will be made by testing multiple models on multiple datasets and data splits. After testing it is found that training directly on IoU significantly increases performance for some models compared to training using the conventional Binary Cross Entropy loss.

Item Type: Thesis (Bachelor's Thesis)
Supervisor name: Wiering, M.A.
Degree programme: Artificial Intelligence
Thesis type: Bachelor's Thesis
Language: English
Date Deposited: 30 Jul 2018
Last Modified: 30 Jul 2018 13:49

Actions (login required)

View Item View Item