Indian Sign Language Recognition Model Development
Budget: ₹600 – ₹1,500 INR
Project Title:
Create a Simple Indian Sign Language Recognition Model for Words, Characters, and Numbers
Project Description:
I need a simple machine learning model that can recognize hand gestures representing Indian Sign Language (ISL). The model should recognize characters (A-Z), numbers (0-9), and some basic words (like hello, thank you, sorry, yes, no, love).
Key Features:
Character Detection (A-Z): Recognize hand gestures for all 26 alphabets.
Number Recognition (0-9): Recognize hand gestures for digits 0-9.
Word Recognition: Recognize common words in ISL (e.g., hello, thank you, sorry, yes, no).
Real-Time Detection: Model should be able to process webcam video in real-time and recognize gestures quickly (under 100ms).
Simple Output: The recognized gestures should be displayed as text and optionally spoken aloud (text-to-speech).
Deliverables:
Trained Model: A trained model that can detect characters, numbers, and basic words in ISL.
Model file format: TensorFlow (.h5) or ONNX (for easy deployment).
Dataset: If you don’t have a pre-existing dataset, guidance on collecting and labeling gestures for A-Z, 0-9, and a small set of words.
Around 100-200 samples per class (character, number, word).
Code for Inference: A simple Python or TensorFlow.js script that can:
Take a live video feed from a webcam,
Recognize the gestures in real-time, and
Output recognized text (and speak it out using gTTS or Web Speech API).
Documentation: A basic README file explaining:
How to run the code,
How to use the trained model,
Simple instructions for setting up and testing.
Skills Required:
Basic Machine Learning experience (TensorFlow / Keras / PyTorch).
Gesture Recognition or Hand Pose Estimation using models like MediaPipe or HandPose.
Real-Time Processing (video or webcam input).
Familiarity with TensorFlow.js (optional, if the model needs to be used in a browser).
Budget:
This is a small project with a limited budget. Please provide an estimate based on basic functionality and a quick turnaround (within 1–2 days).
Timeline:
1-2 Days: We need a working prototype or MVP within 1–2 days.
The goal is to build something functional with basic accuracy for recognizing a few common gestures, not a fully advanced model.
How to Apply:
Please provide:
A brief summary of your experience with gesture recognition or real-time ML models.
Your approach for quickly building the model and achieving real-time recognition.
The estimated cost for the project (be realistic, considering the short timeline).
If available, links to similar projects or any relevant work.
Create a Simple Indian Sign Language Recognition Model for Words, Characters, and Numbers
Project Description:
I need a simple machine learning model that can recognize hand gestures representing Indian Sign Language (ISL). The model should recognize characters (A-Z), numbers (0-9), and some basic words (like hello, thank you, sorry, yes, no, love).
Key Features:
Character Detection (A-Z): Recognize hand gestures for all 26 alphabets.
Number Recognition (0-9): Recognize hand gestures for digits 0-9.
Word Recognition: Recognize common words in ISL (e.g., hello, thank you, sorry, yes, no).
Real-Time Detection: Model should be able to process webcam video in real-time and recognize gestures quickly (under 100ms).
Simple Output: The recognized gestures should be displayed as text and optionally spoken aloud (text-to-speech).
Deliverables:
Trained Model: A trained model that can detect characters, numbers, and basic words in ISL.
Model file format: TensorFlow (.h5) or ONNX (for easy deployment).
Dataset: If you don’t have a pre-existing dataset, guidance on collecting and labeling gestures for A-Z, 0-9, and a small set of words.
Around 100-200 samples per class (character, number, word).
Code for Inference: A simple Python or TensorFlow.js script that can:
Take a live video feed from a webcam,
Recognize the gestures in real-time, and
Output recognized text (and speak it out using gTTS or Web Speech API).
Documentation: A basic README file explaining:
How to run the code,
How to use the trained model,
Simple instructions for setting up and testing.
Skills Required:
Basic Machine Learning experience (TensorFlow / Keras / PyTorch).
Gesture Recognition or Hand Pose Estimation using models like MediaPipe or HandPose.
Real-Time Processing (video or webcam input).
Familiarity with TensorFlow.js (optional, if the model needs to be used in a browser).
Budget:
This is a small project with a limited budget. Please provide an estimate based on basic functionality and a quick turnaround (within 1–2 days).
Timeline:
1-2 Days: We need a working prototype or MVP within 1–2 days.
The goal is to build something functional with basic accuracy for recognizing a few common gestures, not a fully advanced model.
How to Apply:
Please provide:
A brief summary of your experience with gesture recognition or real-time ML models.
Your approach for quickly building the model and achieving real-time recognition.
The estimated cost for the project (be realistic, considering the short timeline).
If available, links to similar projects or any relevant work.