This page contains the most visible part of my projects, including live demos, videos and datasets.
UPDATE: APRIL 2020
Revisionista is currently tracking live Covid-19 related news from 36 publications. Check it here.
Revisionista.PT is a meta-journalistic tool that reveals post-publication changes in articles of Portuguese online news media.
It can uncover the news cycle of online media,
offering a glimpse into an otherwise unknown dynamic edit history.
We leverage on article snapshots periodically collected by Web archives to reconstruct an approximate timeline of the changes: additions, edits, and corrections.
RevisionistaPT is currently tracking changes in about 140,000 articles from the Portuguese Web Archive, published by 12 selected news sources and has a user-friendly interface that will be familiar to users of version control systems.
In addition, an open source browser extension can be installed by users so that they can be alerted of changes to articles they may be reading.
Effortly combines your Strava activity information with your Spotify or last.fm listening habits, enabling you to find the ones that make you go faster and make your heart beat harder.
Moetsi’s Sensor Stream Pipe (SSP) is an open-source C++ API that makes it easier to process both color and depth data by simplifying and expediting remote rendering. It does this by efficiently compressing raw data streams, allowing developers to send multiple video types over the network in real time. Frame data can be sent in its raw form (JPG/PNG frames), or compressed using a myriad of codecs, leveraged on FFmpeg/LibAV and NV Codec to considerably reduce bandwidth strain.
In its current incarnation, Moetsi’s pipe supports the Azure Kinect DK RGB-D camera and existing datasets (e.g. BundleFusion, MS RGB-D 7 scenes and VSFS), but we’re planning to support other cameras and devices (e.g. Kinect v2 and Structure Core) in the near future. At the same time, we’ve made the Moetsi SSP totally open source so that you can build out support for any devices that you want to too!
Nova MedSearch, a medical search engine that integrates the two search modalities: text and image. Our goal is to provide an intuitive and simplified way of supporting multimodal queries in medical search.
Users can upload his own images to build their query or use existing sample images in their queries. The results are displayed in an ranked list with basic information (e.g. title, keywords, images (if available)) and a link to the corresponding article details.
The interface takes into account both the relevancy of the images and text similarity.
This is a game to play with facial expressions without any remote control or gestures – the best actor wins the game! The game dynamics were tuned to increase competitiveness among players and to demonstrate a new affective computation paradigm.
With several machine learning algorithms supporting this game, it integrates an API for image indexing and retrieval. This game demonstrates real-time image analysis and features extraction for search and other applications (as this game for instance). It uploads data into a REST service to index and search images where the SearchMotions application illustrates the potential of the developed search methods. More information about the dataset is available here.