EAST WEST INSTITUTE OF TECHNOLOGY

(Affiliated to Visvesvaraya Technological University, Belgaum, Karnataka)

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

MINI PROJECT SYNOPSIS PRESENTATION
ON
“ GESTURE LANGUAGE RECOGNITION“
Submitted By
PRATIKSHA 1EW21CS119
MADHUSHREE M G 1EW21CS087
KAVYA H 1EW21CS072
NAMEERA ZUHA 1EW21CS103

Under the Guidance of

Prof .SHWETHA N
Designation
Dept of CSE,EWIT

2023-24

06/07/25 3 1
 CONTENTS

•Abstract
•Introduction
•Literature Review
•Problem Statement
•Objectives of the Project
•System Requirement Specifications
•Implementation
•Output
•References (IEEE/Scopus/SCI Indexed papers)

06/07/25 3 2
 ABSTRACT

This project focuses on developing a comprehensive sign language detection system aimed at
bridging communication gaps for the deaf and hard-of-hearing community.
Utilizing advanced computer vision and machine learning techniques, the system is designed to
accurately recognize and translate sign language gestures into spoken or written language in real-
time.
The core of the project involves capturing hand and finger movements through a camera,
preprocessing the input to extract key features, and applying deep learning models to classify the
gestures..
The ultimate goal is to create an accessible, user-friendly application that can be deployed on mobile
and desktop platforms, significantly enhancing communication accessibility and inclusivity.
This project not only aims to improve the quality of life for individuals relying on sign language but
also sets the foundation for future advancements in human-computer interaction and assistive
technologies.

06/07/25 3 3
 INTRODUCTION

• Sign language serves as a vital mode of communication for the deaf and hard-of-
hearing community.
• Despite its importance, there remains a significant communication barrier between
sign language users and those who do not understand it.
• In recent years, advancements in computer vision and machine learning have
paved the way for innovative solutions to bridge this communication gap.
• Sign language detection systems, leveraging these technologies, have the potential
to translate sign language into spoken or written language in real-time, facilitating
seamless communication between signers and non-signers.

06/07/25 Dept of
3 CSE,EWIT 4
 LITERATURE SURVEY

Paper and Author details Description Advantages Disadvantages

Exploring Sign Language Explores various Comprehensive analysis.

-Limited scope.
Detection on Smartphones: methodologies, techniques, -Insightful Evaluation.
A Systematic and challenges encountered -Practical relevance. -Publication Bias.
Review of Machine and in developing robust
Deep Learning Approaches. systems for sign language -Technological
Iftikhar Alam , recognition, offering
obsolescence.
Abdul Hameed, and Riaz insights into the current state
Ahmad Ziar. of the field and avenues for -Inherent complexity.
future research.

06/07/25 3 5
Paper and Author details Description Advantages Disadvantages

Deep sign introduces a novel - Real-time -Complexity and

Deep sign: Sign Language
approach to sign language Processing. Technicality.
detection and recognition detection and recognition -Innovative approach. -Limited Generalization.
Using utilizing deep learning -Deep Learning -Data Requirements.
techniques. By leveraging the Techniques. -Dependency on
Deep Learning. Kothadiya, power of neural networks, it -Scalability. preprocessing

Chintan Bhatt , Krenil aims to accurately detect and

interpret sign language
Sapariya, Kevin Patel, Ana- gestures, offering
Belén Gil-González communication for the deaf and
hard of hearing community
and Juanm M. Corchado.

06/07/25 3 6
Paper and Author details Description Advantages Disadvantages

Sign Language Detection By leveraging convolutional -Improved Accuracy. -Interpretability

using Deep Learning and recurrent neural -Data Privacy
-Robustness.
Smit Patel (1), Tanushree networks, it effectively Concerns.
Pardhi captures both spatial and -Temporal Analysis. -Environmental
(2), Zankhana Shah. temporal features crucial for -Potential Application. Factors.
accurate recognition. The -Limited Dataset.
paper also presents a
specialized dataset tailored
for sign language
detection tasks.

06/07/25 3 7
 PROBLEM STATEMENT

• For the deaf and hard-of-hearing, communication hurdles persist due to

limited accessibility and disruption from traditional methods like notes
or interpreters. Current tech solutions lack accuracy and user-
friendliness, hindering real-time sign language translation. Addressing
this, a dependable, real-time system is imperative, adept at interpreting
diverse signing styles and dialects. Accessible across devices, such a tool
could revolutionize communication, mitigating social isolation and
expanding opportunities in education and employment for sign language
users.

06/07/25 3 8 7
 OBJECTIVES

 Develop a Recognition Model: To develop a machine learning model using TensorFlow that can
accurately recognize and interpret sign language gestures and general hand gestures.

 Implement Real-Time Processing: To utilize OpenCV for capturing and processing video input in
real-time, ensuring that the system can operate efficiently under practical conditions.

 Training and Optimization: Train the CNN model using available datasets, optimizing hyper
parameters to achieve high accuracy and generalization.
 Create an Integrated System: To integrate these technologies into a cohesive system that can
function effectively for recognizing sign language gestures and hand gestures in real-time

3
06/07/25 Dept of CSE,EWIT 9
  Real-Time Implementation: Implement the CNN model to perform inference in real-time scenarios,
ensuring low latency and high efficiency.

 Training and Optimization: Train the CNN model using available datasets, optimizing hyper parameters to
achieve high accuracy and generalization. the CNN model is used to perform inference in real-time
scenarios, ensuring low latency and high efficiency.

06/07/25 3 10
 SYSTEM REQUIREMENTS
1. HARDWARE REQUIREMENTS :
Processor : INTEL hardcore TMi3-7 100u CPU
RAM : 4GB
Hard-disk capacity : 64 bit OS

2. SOFTWARE REQUIREMENTS :
MediaPipe 0.8.1
OpenCV 3.4.2 3
TensorFlow 2.3.0
Matplot.lib
Scikit-learn
IDE used : PyCharm
Algorithm used : CNN
Technology used : Deep Learning

06/07/25 Dept of CSE,EWIT 11

 Implementation
• import csv use_static_image_mode = args.use_static_image_mode
import copy min_detection_confidence = args.min_detection_confidence
import argparse
min_tracking_confidence = args.min_tracking_confidence
import itertools
from collections import Counter use_brect = True
from collections import deque # Camera preparation
cap = cv.VideoCapture(cap_device)
import cv2 as cv cap.set(cv.CAP_PROP_FRAME_WIDTH, cap_width)
import numpy as np cap.set(cv.CAP_PROP_FRAME_HEIGHT, cap_height)
import mediapipe as mp # Model load
from utils import CvFpsCalc
mp_hands = mp.solutions.hands
from model import KeyPointClassifier hands = mp_hands.Hands(
from model import PointHistoryClassifier static_image_mode=use_static_image_mode,
max_num_hands=1,
• def main():
# Argument parsing min_detection_confidence=min_detection_confidence,
args = get_args() min_tracking_confidence=min_tracking_confidence,
cap_device = args.device )
cap_width = args.width keypoint_classifier = KeyPointClassifier()
cap_height = args.height point_history_classifier = PointHistoryClassifier()
06/07/25 3 12
 Output

06/07/25 3 13
 REFERENCES

1. https://www.youtube.com/watch?v=pDXdlXlaCco
2. Article on Exploring Sign Language Detection on Smartphones: A Systematic Review of Machine and Deep Learning Approaches.
Iftikhar Alam , Abdul Hameed, and Riaz Ahmad Ziar.
3. Execution of the project [code and commands] was done on Jupyter and command prompt.
4. Automatic SignWriting Recognition: Combining machine learning and expert knowledge to solve Novel Problem, by ANTONIO F. G
SEVILLA, ALBERTO DIAZ ESTEBAN, AND JOSE MARIA LAHOZ-BENGOECHEA.
5. Hand Sign Language Detection Using Deep Learning BY SHUBHAM SHARMA, APALA GHOSH AND SHARMILA SUBUDHI.
6. https://youtu.be/Ui85SVJsRf8?si=LTzOuUS95B-h8p9H
7. https://youtu.be/EiNyiqx1u2E?si=bCIa7dBQvMlihIta
8. G. B. Altimira, ‘‘The meaning of sign language: Reference, specificity and structure in Catalan sign language discourse,’’ in Sign
Languages and Deaf Communities. Boston, MA, USA: De Gruyter Mouton, 2015.

Dept of CSE,EWIT
06/07/25 3 10