Hand Gesture Recognition Using Deep Learning & CNN – ResNet50

**Project Overview**

Developed a Deep Learning-based Hand Gesture Recognition System using Convolutional Neural Networks (CNN) and Transfer Learning.

The system is designed to automatically recognize and classify hand gestures from images into 10 different categories, demonstrating the application of Computer Vision in human-computer interaction and gesture-based recognition systems.

**Dataset & Preprocessing**

- Worked with a hand gesture dataset containing 10 different gesture classes.

- Applied image resizing and normalization to prepare images for model training.

- Used data augmentation techniques to improve model generalization and reduce overfitting.

- Split the dataset into training, validation, and testing sets.

**Model Development**

Implemented a ResNet50-based image classification model using Transfer Learning.

The development process included:

- Utilizing a pre-trained ResNet50 architecture.

- Adapting the classification layers to recognize 10 hand gesture categories.

- Training and validating the model.

- Evaluating performance using classification metrics.

**Evaluation & Results**

Evaluated the model using:

- Accuracy

- Precision

- Recall

- Confusion Matrix

- Class-wise performance analysis

**Tools & Technologies**

- Python

- PyTorch

- ResNet50

- Convolutional Neural Networks (CNN)

- Transfer Learning

- Deep Learning

- Computer Vision

- Google Colab

**Project Outcome**

Successfully developed an image classification model capable of recognizing different hand gestures, demonstrating practical experience in Deep Learning, Computer Vision, model training, and performance evaluation.