/CV-Nanodegree-Projects

Udacity Computer Vision Nanodegree Projects

Primary LanguageJupyter Notebook

CV-Nanodegree-Projects

Udacity Nanodegree - Computer Vision Projects

This repository contains projects submitted for completing Udacity's Computer Vision Nanodegree.

Projects

  • P1 - Facial Keypoints Detection: Build a facial keypoint detection system.
  • P2 - Image Captioning: Captioning images from COCO dataset
  • P3 - Implementing SLAM: Implementing SLAM algorithm for 2D motion and measurements

Note: The code has been slightly modified for running on local machine with GPU support.

Hardware

  • OS: Windows 10 Pro
  • Processor: Intel Core i7-6700HQ CPU @ 2.60 GHz
  • Ram: 16.0 GB
  • GPU: GeForce GTX 960M

Software

  • Cuda Toolkit v9.2
  • Anaconda

Anaconda Setup

  1. Setup a conda environment with Python 3.6
conda create --name cv-nd-cuda92 python=3.6
activate cv-nd-cuda92
  1. Install Pytorch GPU
conda install pytorch cuda92 -c pytorch
pip3 install torchvision
  1. Install remaining packages
pip install -r requirements.txt