Skip to content

This ASL detection model is built on MobileNetV2, pretrained on ImageNet and fine-tuned on an ASL dataset, optimized for mobile performance, making it ideal for Snap AR.

Notifications You must be signed in to change notification settings

shoryax/ASL-recognition

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

13 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

ASL Recognition

Machine Learning model for recognizing American Sign Language (ASL) letters


πŸ“¦ Dataset

We used a public ASL dataset provided by David Lee on Roboflow, and more containing labeled images for each letter in the ASL alphabet.

  • Format: Object Detection (converted for classification)
  • Classes: A–Z (26 total)
  • Source: Roboflow Public Datasets

🧠 Model

The model is based on MobileNetV2, a lightweight convolutional neural network:

  • βœ… Pretrained on ImageNet
  • πŸ” Fine-tuned on the ASL dataset
  • πŸ”§ Optimized for mobile performance (ideal for Snap AR)

πŸ“Š Accuracy

Metric Accuracy Dataset Size Change from Before Status
Training Accuracy 99.95% 12,789 images Same (was ~100%) βœ… Excellent
Validation Accuracy 97.66% 1,411 images +10.74% πŸš€ βœ… Excellent!
Test Accuracy 87.50% 80 images +0.86% βœ… Very Good

Status: Ready to use! πŸŽ‰

The model generalizes well, maintaining high performance on unseen ASL hand signs.


πŸ§ͺ Image Augmentation

To improve generalization and robustness, the following augmentations were applied during training:

  • πŸ”„ Rotation
  • ↔️ Width & Height Shift
  • πŸ” Zoom
  • πŸ“ Shear Transformation
  • πŸ’‘ Brightness Adjustment

These augmentations simulate real-world variations like lighting and camera angles.


πŸš€ How to Use

  1. Clone this repo:

    git clone https://github.com/your-username/asl-recognition.git
    cd asl-recognition

About

This ASL detection model is built on MobileNetV2, pretrained on ImageNet and fine-tuned on an ASL dataset, optimized for mobile performance, making it ideal for Snap AR.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages