Skip to content

Latest commit

 

History

History
19 lines (16 loc) · 1013 Bytes

README.md

File metadata and controls

19 lines (16 loc) · 1013 Bytes
Raw

Electrothon-3.0-Team-Adway

Project: Sign Language Recognition Using Deep Learning

Initial Commit:

  • We are starting this project with a dataset of single images in each label eg. 1, 2, A, etc.
  • Also the ppt is blank or it is just the template provided.

DataSet Creation

  • After capturing more images and applying data augmentaion, the total dataset now contains more than 10,000 images
  • Images are captured by
    • Vinay
    • Aakansha
  • Dataset is available in my google drive Dataset Link

Creating the CNN Model

  • Model 1: Using data as preprocessed by the tensorflow.keras.applications.mobilenet.preprocess_input but the model is created by us.
  • Model 2: Using the previous model but the data is in grayscale.
  • Model 3: Using the already existing very popular mobilenet model from the tensorflow.keras.applications.mobilenet, and tuning it a little bit as per our requirement. This is our final model.