Peter Onisha Peregbakumo is a BSc Hons Computer Science graduate from our Class of 2019, and also one of the recipients of The Founders’ Award, given to graduating students who best embody the values of Transnational Academic Group (TAG), which are Loyalty, Attitude, Integrity, Competence and Commitment. In Peter’s final year he embarked on a project that made him explore Optical Character Recognition (OCR), Computer Vision and Machine Learning. After graduation, he decided to put out an implementation from his research and work; hence created a unique mobile app called PaperLyft.
Print media has been around for many years and has only evolved to digital form because of the invention of more technological devices that allow us to share contents through social media and even through office software for workers. It’s no secret that printed documents have always been commonly found in the office, with companies developing various filing methods to store important documents. Unfortunately, users often don’t get access to the soft copy of the documents and if they really need the information from the printed paper or item, they would have to manually type it out into their PCs or mobile phones, or scan documents and convert into an editable format.
The stress that this brings varies depending on the length of the document, and the amount of content that needs to be typed. Peter experienced a real life example of a case where an employee was given 500 business cards to type the details of each card unto an excel sheet. The time used to do this would have been used to do more productive tasks like actually composing emails for the email list assuming the contents of the card could “magically” appear on the excel sheet without going through the stress of typing of every single detail. This problem and other similar problems that workers and students face inspired the creation of “PaperLyft”, your one stop for extracting text from any printed media including and not limited to documents, newspapers, business cards, vouchers, receipts, bank issued cards (it’s a criminal offense to use this on someone else’s card), billboards, flyers etc.
This revolutionizes the whole experience with manipulating printed content within a shorter period of time, and with much less effort. Taking pictures is a simple activity that everyone is either used to, and in extreme cases, even addicted to. Almost everyone has a smartphone, and they are commonly used for social media and watching videos or taking pictures. With PaperLyft, your phone can be worth more to you.
PaperLyft uses computer vision and machine learning with trained data to detect English, French, Portuguese and Spanish characters in the exact way it’s presented on the photo source. This allows the user to copy the extracted text in the same format to his/her preferred text editor or work environment and have the freedom of editing or adding on to existing work.
We are extremely proud of Peter, and look forward to seeing and hearing of his future accomplishments!
PaperLyft’s Android version is available for download currently and if you want to know more about the creator or the app, visit the official PaperLyft website.