Back to all projects
Deep Learning

Object Recognition with Pre-trained Models

Utilizing pre-trained deep learning models for efficient object recognition tasks.

View on GitHub
Object Recognition with Pre-trained Models screenshot

Project Overview

A project demonstrating the power of transfer learning by using state-of-the-art pre-trained Convolutional Neural Networks (CNNs) for various object recognition tasks.

Technologies Used

PythonTensorFlowKerasTransfer Learning

Methodology

  1. 1Loading pre-trained models (e.g., VGG16, ResNet50, MobileNetV2).
  2. 2Fine-tuning the last layers for specific datasets and classes.
  3. 3Performing inference on new images and videos for object detection.
  4. 4Understanding feature extraction from deep layers and model architectures.

More in Deep Learning

Covid-19 Detection from X-Rays

Using Deep Learning (CNN) to detect Covid-19 from chest X-ray images.

Web Ad Optimization (RL)

Solving the Multi-Armed Bandit problem for advertisement clicks using Upper Confidence Bound.

Diabetes Risk Prediction

Predicting diabetes onset in high-risk populations using Neural Networks and diagnostic data.