https://github.com/shoaib-33/fighter-jet-image-recogniser

Link to Git Page
https://github.com/shoaib-33/fighter-jet-image-recogniser

computer-vision deep-learning hugging machine-learning

Last synced: about 1 month ago
JSON representation

Link to Git Page

• Host: GitHub
• URL: https://github.com/shoaib-33/fighter-jet-image-recogniser
• Owner: Shoaib-33
• License: mit
• Created: 2023-07-07T13:26:56.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-07-29T16:57:01.000Z (over 1 year ago)
• Last Synced: 2023-07-29T17:28:25.788Z (over 1 year ago)
• Topics: computer-vision, deep-learning, hugging, machine-learning
• Language: Jupyter Notebook
• Homepage: https://shoaib-33.github.io/Fighter-Jet-Image-Recogniser/
• Size: 99.5 MB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Fighter-Jet-Image-Recognizer



# Description

The project is all about Military aviation more specifically fighter jet image recoginizer which can identify Military aircraft from the image of the aircraft provided as input.

# Technical Specification:

1. FastAi for data preprocessing,training etc.

2. Gradio.

3. Hugging face space for deployment.

4. GitHub page for API

# Fighter-Jet-Image-Recognizer

An image classification model from data collection, cleaning, model training, deployment and API integration. 


The model can classify 26 different types of Fighter Aircrafts 


The types are following: 


1. General Dynamics F-16 Fighting Falcon aircraft

2. Mikoyan MiG-29

3. Dassault Rafale

4. Lockheed Martin F-22 Raptor

5. Mikoyan Mig-31

6. Sukhoi Su-27

7. Lockheed Martin F-35 Lightning II

8. KAI KF-21 Boramae

9. Sukhoi Su-57

10. Shenyang FC-31 Gyrfalcon

11. Boeing F-15EX Eagle II

12. Eurofighter Typhoon

13. AIDC F-CK-1 Ching-kuo

14. Lockheed YF-12

15. Mikoyan MiG-31

16. McDonnell Douglas F-4 Phantom II

17. Grumman F-14 Tomcat

18. English Electric Lightning

19. Focke-Wulf Fw 190

20. Mikoyan-Gurevich MiG-25

21. Lockheed P-80 Shooting Star

22. Sukhoi Su-34

23. Chengdu J-20

24. Sukhoi Su-35

25. Chengdu J-10

26. Mikoyan MiG-35

# Dataset Preparation

**Data Collection:** Downloaded from DuckDuckGo using term name 


**DataLoader:** Used fastai DataBlock API to set up the DataLoader. 


**Data Augmentation:** fastai provides default data augmentation which operates in GPU. 


Details can be found in `notebooks/data_prep.ipynb`

# Training and Data Cleaning

**Training:** Fine-tuned a resnet34 model for 25 epochs (8 times) and got upto ~84% accuracy. 


**Data Cleaning:** This part took the highest time. Since I collected data from browser, there were many noises. Also, there were images that contained. I cleaned and updated data using fastai ImageClassifierCleaner. I cleaned the data each time after training or finetuning, except for the last time which was the final iteration of the model. 


# Model Deployment

I deployed to model to HuggingFace Spaces Gradio App. The implementation can be found in `deployment` folder or [here](https://huggingface.co/spaces/Shoaib-33/Fighter-jet-Image-Recogniser). 




# API integration with GitHub Pages

The deployed model API is integrated [here](https://shoaib-33.github.io/Fighter-Jet-Image-Recogniser/) in GitHub Pages Website. Implementation and other details can be found in `docs` folder.