|
|
--- |
|
|
title: Image Classification with CNN |
|
|
emoji: 🔥 |
|
|
colorFrom: yellow |
|
|
colorTo: green |
|
|
sdk: docker |
|
|
pinned: false |
|
|
--- |
|
|
|
|
|
|
|
|
# Convolutionnal Neural Network Model for Image CLassification Classification |
|
|
|
|
|
## Model Description |
|
|
|
|
|
This is aCNN model for the Frugal AI Challenge 2024, specifically for the image classification task of identifying smoke in images. The model contains 2 convolutionnal layers and one fully connected layer. |
|
|
|
|
|
### Intended Use |
|
|
|
|
|
- **Primary intended uses**: Test for image classification models |
|
|
- **Primary intended users**: Researchers and developers participating in the Frugal AI Challenge |
|
|
- **Out-of-scope use cases**: Not intended for production use or real-world classification tasks |
|
|
|
|
|
## Training Data |
|
|
|
|
|
The model uses the pyronear/pyro-sdis datase. |
|
|
The Pyro-SDIS Subset contains 33,636 images, including: |
|
|
- 28,103 images with smoke |
|
|
- 31,975 smoke instances |
|
|
- Split: 80% train, 20% test |
|
|
|
|
|
|
|
|
## Performance |
|
|
|
|
|
### Metrics |
|
|
- **Accuracy**: ~83% |
|
|
- **Environmental Impact**: |
|
|
- Emissions tracked in gCO2eq |
|
|
- Energy consumption tracked in Wh |
|
|
|
|
|
### Model Architecture |
|
|
The model implements a CNN model trained on augmented images (randomCrop, Horizontal and Vertical Flip, ColorJitters...). Only 2 convolutionnal layers and one fully connected layer was implemented in this model. |
|
|
|
|
|
## Environmental Impact |
|
|
|
|
|
Environmental impact is tracked using CodeCarbon, measuring: |
|
|
- Carbon emissions during inference |
|
|
- Energy consumption during inference |
|
|
|
|
|
This tracking helps establish a baseline for the environmental impact of model deployment and inference. |
|
|
|
|
|
## Limitations |
|
|
- No object detection |
|
|
- |
|
|
|
|
|
## Ethical Considerations |
|
|
|
|
|
- Dataset contains sensitive topics related to climate disinformation |
|
|
- Model makes random predictions and should not be used for actual classification |
|
|
- Environmental impact is tracked to promote awareness of AI's carbon footprint |
|
|
``` |
|
|
|
