Scribd
Upload
46 views

Sign Language 1.1

This document summarizes a dual-camera vision system for sign language recognition using convolutional neural networks. The system uses two cameras, one mounted on the head and one on the chest, to obtain different views of each sign for identification. A dataset of 24 static Panamanian sign language letters was used to train an Inception V3 image classification model in TensorFlow to classify the images. The project demonstrates a prototype for sign language recognition using a first-person dual-camera setup and deep learning techniques.

Uploaded by

NIYATI KARANI
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
46 views

Sign Language 1.1

This document summarizes a dual-camera vision system for sign language recognition using convolutional neural networks. The system uses two cameras, one mounted on the head and one on the chest, to obtain different views of each sign for identification. A dataset of 24 static Panamanian sign language letters was used to train an Inception V3 image classification model in TensorFlow to classify the images. The project demonstrates a prototype for sign language recognition using a first-person dual-camera setup and deep learning techniques.

Uploaded by

NIYATI KARANI
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 8

SIGN LANGUAGE

ALPHABET
RECOGNITION
TEAM PRESENTATING
MEET KINJAL
BHANUSHALI  VAGHASIYA 
 TE A EXTC TE B EXTC
ROLL NO 05 ROLL NO 60

NIYATI KARANI  MEET BHANSALI 


TE B EXTC  TE A EXTC 
 ROLL NO 01 ROLL NO 04
INTRODUCTION
• Sign language recognition is a problem that has been addressed in
research for years. However, we are still far from finding a complete
solution available in our society.
• Among the works developed to address this problem, the majority
of them have been based on basically two approaches: contact-
based systems, such as sensor gloves; or vision-based systems,
using only cameras. The latter is way cheaper and the boom of deep
learning makes it more appealing.
• This post presents a prototype of a dual-cam first-person vision
translation system for sign language using convolutional neural
networks. This is divided into three main parts: the system design,
the dataset, and the deep learning model training and evaluation.
VISION SYSTEM

Vision is a key factor in sign language, Trying to understand sign The vision system is
and every sign language is intended to language from a first-vision composed of two cameras: a
be understood by one person located in perspective has the same head-mounted camera and a
front of the other, from this perspective, limitations, some gestures will chest-mounted camera. With
a gesture can be completely observable. end up looking the same way. these two cameras we obtain
Viewing a gesture from another But, this ambiguity can be solved two different views of a sign,
perspective makes it difficult or almost by locating more cameras in a top-view, and a bottom-
impossible to be understood since every different positions. In this way, view, that works together to
finger position and movement will not what a camera can’t see, can be identify signs.
be observable. perfectly observable by another
camera.
DATASET

To develop the first prototype of this system


is was used a dataset of 24 static signs from
the Panamanian Manual Alphabet. To model
this problem as an image recognition
problem, dynamic gestures such as letter J, Z,
RR, and Ñ were discarded because of the
extra complexity they add to the solution.
IMAGE CLASSIFICATION
TENSERFLOW

This model has been pre-trained for


A generic image classification
the ImageNet Large Visual
program that uses Google's Recognition Challenge using the data
Machine Learning
from 2012, and it can differentiate
library, Tensorflow and a pre- between 1,000 different classes, like
trained Deep Learning
Dalmatian, dishwasher etc. The
Convolutional Neural Network
program applies Transfer Learning to
model called Inception.
this existing model and re-trains it to
classify a new set of images.
REQUIREMENT
S
This project uses python 3.5 and the PIP following packages:
• opencv
• tensorflow
• matplotlib
• numpy
CONCLUSION AND
REFERENCE
• This project is a sign language alphabet recognizer using Python, openCV
and tensorflow for training InceptionV3 model, a convolutional neural
network model for classification.
• The framework used for the CNN implementation can be found here:
• Simple transfer learning with an Inception V3 architecture model by
xuetsing
• The project contains the dataset (1Go). If you are only interested in code,
you better copy/paste the few files than cloning the entire project.

You might also like