https://github.com/clementsan/cnn-image-classification-uncertainty

AI-based image classification tasks with uncertainty estimation
https://github.com/clementsan/cnn-image-classification-uncertainty

cnn-model deep-learning fastai image-classification

Last synced: 6 months ago
JSON representation

AI-based image classification tasks with uncertainty estimation

• Host: GitHub
• URL: https://github.com/clementsan/cnn-image-classification-uncertainty
• Owner: clementsan
• Created: 2023-12-14T21:18:35.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2023-12-14T21:32:13.000Z (almost 2 years ago)
• Last Synced: 2025-02-04T13:03:32.850Z (8 months ago)
• Topics: cnn-model, deep-learning, fastai, image-classification
• Language: Jupyter Notebook
• Homepage:
• Size: 22.5 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# AI-based image classification tasks with uncertainty

Image classification tasks via CNN models, leveraging Monte Carlo dropout during inference to assess model uncertainty

Related paper: [Gal et al., [Dropout as a bayesian estimation: representing model uncertainty in deep learning](http://proceedings.mlr.press/v48/gal16.pdf), ICML 2016

## Analysis tasks

### Image preprocessing

  - Generate small image tiles from larger acquisitions

  - Split dataset into training, validation and test sets, using stratification strategy at acquisition level

### Image analysis

  - Neural network training

  - Neural network inference (direct)

  - Neural network inference with gradCAM visualization

  - Neural network inference with uncertainty assessment

## Notes

### Neural networks

Use of ResNet models

### Deep learning library

Use of fastai_v1 (2019)