This project explores various deep learning architectures for object recognition. It's structured into three main approaches: Feedforward Neural Networks (ffNN), Convolutional Neural Networks (CNN), and Transfer Learning (tLearning). Each method is designed to evaluate different strategies and improve the accuracy of object recognition.
- Deep_Learning___Report.pdf: Contains the comprehensive report detailing the methodologies, experiments, results, and analysis of the object recognition project.
- DeepLearningAssignment.pdf: Provides the project assignment and objectives that were addressed throughout the course of this research.
- Architectures: Directory containing different model architectures used in the project:
- ffNN: Implementations and experiments using feedforward neural networks.
- CNN: Notebooks and resources related to convolutional neural networks.
- tLearning: Application of transfer learning methods.
To get started with this project, clone this repository and explore the Jupyter notebooks contained within each architecture's folder. The notebooks are self-contained and include comments explaining each step of the process.
Ensure you have Python installed along with the following libraries:
- TensorFlow
- Keras
- NumPy
- Matplotlib
You can install the necessary libraries using pip:
pip install tensorflow keras numpy matplotlib Running the Notebooks To run the notebooks, navigate to the specific architecture directory in your terminal and launch Jupyter Notebook:
cd path_to_directory jupyter notebook
DeepLearningProject/ │ ├── Deep_Learning___Report.pdf ├── DeepLearningAssignment.pdf │ └── Architectures/ ├── ffNN/ │ └── ... (Jupyter notebooks and resources for ffNN) ├── CNN/ │ └── ... (Jupyter notebooks and resources for CNN) └── tLearning/ └── ... (Jupyter notebooks and resources for Transfer Learning)
- Federico Paschetta
- Cecilia Peccolo
This project is licensed under the MIT License.
Special thanks to Universidad Politécnica de Madrid for guidance and resources throughout the project.