• Sabina Pokhrel

BeWithMeReader - Innovation with Xailient Technology

We are at a time where AI technology is within reach of every innovator. Like every new technology, it has opened countless possibilities. BeWithMeReader is one such recent innovation.

Matthew Gifford, a grandfather wanting to be connected to his grandchildren and be with them during their learning, even when not physically present, led to the innovation of the BeWithMeReader.

BeWithMeReader powered by Xailient technology
BeWithMeReader powered by Xailient technology

Matthew developed a new device to read any children’s books to a child in their parent’s or grandparent’s voice. It teaches them how to read and helps them recognise the book’s words and learn how to pronounce the words.

BeWithMeReader is a low-cost device, powered by Xailient Technology, that can be clipped to any book, allowing children to enjoy the book at their own pace while still having their parents or grandparents read to them.

It comprises a camera to take images, a microphone to listen to the child speaking, a speaker to play the audio, and an optional monitor to play video of the reader.

BeWithMeReader uses Xailient AI technology to identify which book the child is reading and the current page number the child is looking at.

See it in action

How does it work

BeWithMeReader is intuitive for a child to use.

  • The child turns the device on with a switch.

  • The child clips book onto a clipboard with the clip over the back cover.

  • The child turns to the front or back flap so the system can ID book/Recording voice.

  • The child turns to any page, and the system should read the page with a selected voice.

  • The system turns off automatically after 5 min of no page turns.

BeWithMeReader detecting page sticker on book using Xailient Detector
BeWithMeReader detecting page sticker on book using Xailient Detector

What happens behind the scenes?

When preparing a book, the teacher, which can be parents or grandparents, can print stickers and place them on existing books as a bookID. On each book page, a sticker with the page number is placed on top of the book. The teacher records the audio and or video for the book and loads it to BeWithMeReader.

When the book is clipped onto the clipboard with back cover, Xailient Detector first detects the bookID sticker, and then Xailient OCR reads the book id. BeWithMeReader then looks up the recording for the book and gets ready to play.

When the child turns to any page, Xailient Detector detects the sticker on the page, and Xailient OCR reads the page number from the detected sticker. The device then plays the associated recording for that page.

Sticker detection and OCR reading are done on the device itself. Since both the sticker detection and reading are done on the device, there’s no dependency on the internet for normal operation.

BeWithMeReader is portable and battery-operated. Best of all, it processes fast enough so children don’t get bored waiting.

Xailient technology fits perfectly with the requirements of the BeWithMeReader application as it is optimized for the edge device, consumes low power, and can operate without an internet connection.

How did he do this?

Matthew collected images of stickers for training the Xailient Detector and the OCR Reader. He used Xailient Console to train the Sticker Detector AI model and an OCR model to read the stickers. He then deployed the Xailient Detector and OCR Reader to his Raspberry Pi device.

“I have created a very inexpensive system that bridges the gap between hard textbooks and e-textbooks with multi-media content. “ - Matthew

To learn more about BeWithMeReader, contact reader@lexride.com

Register for a free account and train your AI model now 👈

Looking to implement real-time face detection on a Raspberry Pi? Check out this post.

More stories:

Integrate Face Detection on your App

Challenges of Running Deep Learning Computer Vision on Computationally Limited Devices

Struggles of Running Object Detection on a Raspberry Pi

You Need to Move from Cloud Computing to Edge Computing Now!

85 views0 comments