Introduction to Visual Question Answering: Datasets, Approaches and Evaluation
Historically, building a system that can answer natural language questions about any image has been considered a very ambitious goal. Imagine a system that, given the image below, could answer these questions: What is in the image? Are there any humans? What sport is being played? Who has the ball? How many players are in the image? Who are the teams? Is it raining? Argentina facing England in 1986Image source
Faster R-CNN: Down the rabbit hole of modern object detection
Previously, we talked about object detection, what it is and how it has been recently tackled using deep learning. If you haven’t read our previous blog post, we suggest you take a look at it before continuing. In 2017, we decided to get into Faster R-CNN, reading the original paper, and all the referenced papers (and so on and on) until we got a clear understanding of how it works and how to implement it.
Top 10 Python libraries of 2017
Update (12/28/2018): other editions of this post: 2018, 2016, 2015. December is the time when you sit back and think about the accomplishments of the past year. For us programmers, this is often looking at the open source libraries that were either released this year (or close enough), or whose popularity has recently boomed because they are simply great tools to solve a particular problem. For the past two years, we have done this in the form of a blog post with what we consider to be some of the best work that has been done in the Python community.
Deep Learning for NLP: Advancements & Trends
Over the past few years, Deep Learning (DL) architectures and algorithms have made impressive advances in fields such as image recognition and speech processing. Their application to Natural Language Processing (NLP) was less impressive at first, but has now proven to make significant contributions, yielding state-of-the-art results for some common NLP tasks. Named entity recognition (NER), part of speech (POS) tagging or sentiment analysis are some of the problems where neural network models have outperformed traditional approaches.
TryoSummerParty, 2017 edition
It’s that time of the year again! Time for uncomfortable family meetings, TODO lists that will be quickly forgotten, crazy last minute holiday shopping and, of course, a new edition of the famous #TryoSummerParty. An as-incomplete-as-you-can-get photo of the TryoGang at this year's #TryoSummerParty. If there’s one thing at which we are almost as good as coding, it’s definitely parties and celebrations. So, as we do every year, last weekend we all got together along with our significant others and kids (yes, we are growing up) to celebrate another amazing year at Tryolabs and also look back at all the great things we accomplished in 2017.
Launching Requestium: An integration layer between Requests and Selenium for automation of web actions
Our ODSC talks in video
As we previously announced, the last couple of weeks we’ve been launching Luminoth, our brand new open source toolkit for Computer Vision. As part of this launch, we gave talks in the Open Data Science Conference (ODSC), both at the London and San Francisco editions. The events were a great success, as in 5 days they accumulated around 5000 data scientists, while hosting more than 200 talks from renowned people in the academia and industry.
And the Tryolabs Scholarship goes to...
Earlier this year we announced the creation of the Tryolabs Scholarship, aimed to support university students at early stages of their careers in computer science and engineering. The experience as a whole was amazing and we selected a winner and a runner-up a couple of months ago. Now that things settled down, we are ready to tell you more about them. And Tryolabs Scholarship goes to… (drum roll please) Let’s start by introducing two outstanding people, examples of hard work, dedication and determination.
Launching Luminoth: our open source computer vision toolkit
After a few months working in stealth mode, we are very proud to launch our Deep Learning initiative: luminoth.ai Luminoth is an open source toolkit for computer vision. Currently, we support object detection and image classification, but we are aiming for much more. It is built in Python, using Google’s Machine Intelligence framework, TensorFlow; and Sonnet, a very useful library built by DeepMind for building complex neural networks with reusable components.
Tryolabs Working Trip NYC, in pictures
Our commitment at Tryolabs is to create the best hi-tech company to be part of. We believe this is done by creating the best opportunities in terms of professional and personal growth for our team. This year we decided to do our first “Tryolabs working trip”. We rented a house in Brooklyn for one month for our team to visit and work from there. Read more about it here. The idea in a nutshell: