Teaching machines to think, reason and communicate.

Microsoft Research Maluuba, Montreal

iStock_85503939_XXXLARGE (1).jpg

Introducing FigureQA: the first annotated question-answering dataset on bar, line, and pie charts.

Answering questions about a given image is a difficult task, requiring both an understanding of the image and the accompanying query. Maluuba's FigureQA dataset introduces a new visual reasoning task for research, specific to graphical plots and figures. The task comes with an additional twist: all of the questions are relational, requiring the comparison of several or all elements of the underlying plot.

Images are comprised on five types of figures commonly found in analytical documents. Fifteen question types were selected for the dataset concerning quantitative attributes in relational global and one-vs-one contexts. These include properties like minimum and maximum, greater and less than, medians, curve roughness, and area under the curve (AUC). All questions in the training and validation sets have either a yes or no answer.

learn more


Dedicated to tackling big challenges in language understanding and artificial intelligence

With a focus on deep learning and reinforcement learning, our growing team of renowned experts is working closely with industry and academia. Now, as part of Microsoft, Maluuba is accelerating our approach to drive breakthroughs in AI research and application. Learn more about our vision, access our academic publications and interact with our public datasets.





Working at Maluuba

"Our research seeks to create literate machines. It’s an ambitious goal and we need big thinkers and dreamers to help achieve it."

Kaheer Suleman, CTO and Co-Founder

We're looking for talented and motivated people to join our team!  Learn more about our vision, our work culture and benefits  and view our career opportunities.

Careers BG.png