Computer Vision Systems | Nikhilesh Narkhede

Project Overview

Designed and implemented multiple CNN-based computer vision pipelines for image classification, object detection, and facial recognition
Built face detection and recognition systems using MTCNN and VGGFace embeddings to identify and verify human faces
Applied transfer learning techniques by fine-tuning ImageNet-pretrained models to improve accuracy on limited datasets
Developed deep CNN models for handwritten digit classification (MNIST) and medical image diagnosis using chest X-ray images
Preprocessed and augmented image datasets to enhance model robustness and generalization
Evaluated model performance across multiple vision tasks, emphasizing accuracy, generalization, and inference reliability

Face detection and recognition system using MTCNN and VGGFace embeddings to identify and verify human faces in images.

Transfer learning model fine-tuning ImageNet-pretrained architectures to achieve high accuracy on custom image classification tasks.

Deep convolutional neural network for handwritten digit classification achieving high accuracy on the MNIST dataset.

Medical image diagnosis system using CNN to detect pneumonia from chest X-ray images with high classification accuracy.

Python Computer Vision Deep Learning CNNs Transfer Learning ImageNet MTCNN VGGFace Medical Imaging TensorFlow/Keras