*Please note the change of date to the 13th November*
This month, we'll learn to use TensorFlow (https://www.tensorflow.org/), a library for machine learning that can be used for image recognition. By the end, we'll have a Python script that 'recognises' an object in an image and classifies it accordingly.
The interactive news team at the New York Times built an image recognition service (https://open.nytimes.com/how-the-new-york-times-uses-software-to-recognize-members-of-congress-29b46dd426c7) that receives photos via SMS and replies with the names of any members of Congress in the image.
Sky News used a similar approach (https://www.theverge.com/2018/5/4/17318354/royal-wedding-uk-facial-recognition-sky-news-date) to automate the identification of attendees during the recent royal wedding.
The recent rapid pace of developments in machine learning has led to powerful tools like TensorFlow becoming more accessible to non-engineers. Journalists are just beginning to explore the newsroom applications of computer vision technology.
Bring a laptop along as this is a practical, hands-on workshop. No programming experience is required. However, if you feel comfortable doing so, please ensure you have Python 3.5 installed and download TensorFlow (https://www.tensorflow.org/install/). If you have trouble doing that, don't worry, we will help you on the day.
Please make sure you're signed up to Dropbox to view our shared doc (http://bit.ly/journocoders-nov-2018) — and add links to anything data or journalism-related you've seen or worked on to the show and tell section!
7:00 🚪 Doors open
7:30 🗣 Show and tell
7:40 💻 Tutorial
9:00 🍺 Drinks at the George
What is Journocoders?
Journocoders is a monthly meetup for journalists and others working in the media to learn and share technical skills for use in their reporting. We aim to bring the culture of knowledge sharing from the tech industry to journalism and give people the chance to learn and network in a supportive environment with likeminded individuals.