File size: 2,774 Bytes
31f9995
b33e5d5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
31f9995
 
 
 
 
 
 
b33e5d5
31f9995
b33e5d5
31f9995
b33e5d5
31f9995
b33e5d5
31f9995
b33e5d5
31f9995
 
 
 
 
 
 
 
 
 
b33e5d5
31f9995
 
 
b33e5d5
 
 
31f9995
b33e5d5
 
 
 
31f9995
b33e5d5
 
 
 
 
 
31f9995
b33e5d5
31f9995
b33e5d5
 
 
31f9995
b33e5d5
 
 
 
 
31f9995
b33e5d5
 
 
 
31f9995
b33e5d5
31f9995
b33e5d5
 
 
31f9995
b33e5d5
 
 
 
 
 
 
 
 
 
 
 
 
 
31f9995
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
---
pretty_name: "Deep Learning Project"
language:
  - en
license: other
task_categories:
  - image-classification
tags:
  - deep-learning
  - computer-vision
  - fire-detection
  - wildfire-detection
  - image-classification
  - transfer-learning
  - severity-classification
  - feature-extraction
  - recommendation-generation
  - tensorflow
  - keras
---

# Deep Learning Project

## Dataset Summary

This repository contains the datasets, trained models, notebooks, experiments, feature-extraction outputs, and supporting resources developed for a deep learning project focused on **fire detection, fire severity classification, and related computer vision tasks**.

The project covers multiple stages of a deep learning workflow, including binary fire classification, three-class fire severity classification, feature extraction, dimensionality reduction, clustering, and recommendation generation.

The repository contains approximately **2.1 GB of files across 1,700+ files**.

## Dataset Details

### Dataset Description

The repository is a collection of datasets and machine-learning artifacts rather than a single standardized dataset. It contains resources used across multiple deep learning experiments and application components.

The main components include:

- Fire vs. No-Fire binary image classification
- Three-class fire severity classification
- Feature extraction
- Severity clustering
- Dimensionality reduction
- Recommendation generation
- Generated severity-image samples
- Jupyter notebooks
- Trained models and model checkpoints
- Supporting application resources

### Main Project Components

#### Fire vs. No-Fire Binary Classification

`Fire_vs_No_Fire_Binary_Classification/`

Contains resources for binary image classification between:

- Fire
- No Fire

The project includes experiments using:

- ResNet50
- Custom CNN
- VGG16
- EfficientNetB0

The directory also contains a dataset, trained model resources, and VGG16 checkpoints.

#### Fire Severity Detection

`Severity_Detection_Tri_Classification/`

Contains resources for three-class fire severity classification:

- Mild
- Moderate
- Severe

The project includes experiments using:

- Xception
- EfficientNetB0

Additional components include feature extraction, clustering, dimensionality reduction, and dataset preparation.

#### Severity Dataset

`Severity_Detection_Tri_Classification/Severity_Altered_Dataset/`

Contains an image dataset organized into training, validation, and testing splits.

```text
severity_dataset/
├── train/
│   ├── mild/
│   ├── moderate/
│   └── severe/
├── val/
│   ├── mild/
│   ├── moderate/
│   └── severe/
└── test/
    ├── mild/
    ├── moderate/
    └── severe/