Zyada

Zyada is an object detection mobile application with augmented reality, that uses a computer vision model to recognize exhibits in a museum, and give the visitors a more enhancing visitor experience.

Many museum-goers find it challenging to fully grasp and remember exhibits due to unengaging content, leading to a diminished learning experience. Addressing this, a cloud backed mobile app utilizing state-of-the-art technology was developed to boost visitor engagement and understanding. Featuring advanced real-time object detection, immersive 3D models, and augmented reality (AR), the app enriches the visitor experience by providing in-depth, interactive information on exhibits. Developed in collaboration with the Masterpiece Museum, this project harnesses the latest technological advancements to significantly enhance the educational impact of museum visits.

TEch used
Flutter, AWS Amplify, Amazon S3, DynamoDB, TensorFlow, Google modelviewer, Python, Computer Vision tools
Date
June 2023
Recognition
Top 12 IT projects in Bahrain from HH SH Khalid Award for Graduation Projects
Affiliation
University of Bahrain

The project has three main components; the application, the cloud backend and the computer vision model.

  • Application: The application was made using flutter framework aiming to make it cross platform, and the need to be as native as possible for a seemless real-time experience. Many packages were used including model-viewer, which eased the AR and 3D feature.
  • Cloud: AWS Amplify was chosen as the backend where the data is stored in a NoSQL DynamoDB database, and 3D models, images stored in Amazon S3. AWS Cognito was used for user authentication and data was queried using GraphQL
  • Computer Vision Model: A CNN model was trained on a custom dataset of over 27,000 images, collected from the Masterpiece Museum and was trained on pretrained weights of YOLOv8n. The resulting model gave an mean Accuracy and Precision score of 99.3%, which resulted in a very reliable model to be used. The model was then optimized for mobile use using TensorFlow.

Presenting Zyada to Sh.Khalid bin Hamad Alkhalifa

Zyada was selected as Top 12 IT projects in Bahrain for the prestigious Khalid bin Hamad Award for Graduation Projects

Exhibit detection

Our custom computer vision model in work!

AR Feature

Had our fun throughout the development process

Testing at MAsterpiece Museum

real-time and accurate responses

Detailed Information

More about the exhibit in text, images, audio, video and 3D models

AWS Cloud Architecture

Compact design optimized for scalability

Nullam quis risus eget urna mollis ornare vel eu leo. Aenean lacinia bibendum nulla sed 

  • @Qreato
Facebook Twitter Youtube Instagram Linkedin