Skip to content

Clothing segmentation and segment classification using CNNs for Udacity capstone

Notifications You must be signed in to change notification settings

magee256/clothing_object_recognition

Repository files navigation

Machine Learning Engineer Nanodegree

Specializations

Project: Capstone Proposal and Capstone Project

Note

The Capstone is a two-staged project. The first is the proposal component, where you can receive valuable feedback about your project idea, design, and proposed solution. This must be completed prior to your implementation and submitting for the capstone project.

You can find the capstone proposal rubric here, and the capstone project rubric here. Please ensure that you are following directions correctly before submitting these two stages which encapsulate your capstone.

You will find an open_projects folder in these files. This will be where pre-curated capstone projects are available, should you choose to work on a project already partially designed for you.

Please email machine-support@udacity.com if you have any questions.

Necessary Files

Instructions on how to download the necessary data and pretrained models follows:

DeepFashion dataset:

Navigate to this URL and download the Clothes Images zip file, bounding box annotations, category annotations and Train/Val/Test partitions.

Pretrained Bounding Box Predictor:

Follow the Getting Started instructions in this Github repo

Pretrained ImageNet Models:

These may be imported from the keras.applications module.

Segmentation Algorithms:

These can be found in the skimage library and a demonstration of them being used can be found here

About

Clothing segmentation and segment classification using CNNs for Udacity capstone

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published