Research


Newslinn started as a research project in 2014 and is continuing to invest time in photo analysis, image verification and classification. Below is an overview of our existing research projects, the current status and if it's open or closed for external participation.

For each research project we are making anonymous data sets available and making our source code available under the MIT License or Apache License v2.0 where possible.

You can contact us to express interest or track our projects on kaggle.

A short note on technology

Newslinn is a technology company. Our mission is to build the technology to help people around the World (in both low-tech and high-tech countries) to communicate easily and safely with journalists - and by doing so provide that technology for free as a platform for journalists. You can read more about our technology here.

Join the conversation

If you wish to participate please join our open Slack channel #newslinn or you can if you prefer.

Project: Image Validation

Patent Applied for S20150216, Digital Image Validation

As part of our image validation patent technology (Patent Applied S20150216), we are looking for collaborators and will be releasing the research and implementation under the MIT License. Once the patent is pending we will be opening this project for external collaboration.

Irish Patent Office Journal

Project: Photo Classification

Part of Enterprise Ireland Innovation Vouchers Scheme and open for Knowledge Providers in Ireland.

As part of our extension to our artificial intelligence rule engine we are researching on supervised classification for image fraud analysis for the prediction of acceptance of photos.

The purpose is to use supervised classification to complement the existing rule engine and to predict if a photo should be 'accepted' or 'declined'.

The project will last 6 months and we will be releasing an anonymized data set that will be open to external parties.

* please note, no photos will be included in the data set.

Project: Photo Error Level Analysis

Internal Research, open to external source code collaboration.

As part of our real-time fraud detection system we are conducting research into different ELA algorithms on a real-time photo data set.

The end outcome will be a research paper on the different approaches and algorithms tests with a particular focus on real-time scalable and reliable algorithms.

Our implementation will be using Python and source code will be released under the MIT License.

Project: Verified Anonymous Sharing

Internal Research, open to external source code collaboration.

As part of our vision we are researching tooling to provide anonymous validated image sharing from sources. This will involve image encryption, steganography, user anonymization and combined our existing efforts on image verification.

This project will be open-sourced with it's aim to be used by third party networks and media companies.