Projects | Bruno Marques

Deep Learning Applied to Optical Physics.

Mon, 27 Apr 2020 00:00:00 +0000

The recent progress in Deep Learning methods has been contributing to a wide range of domains. This project focus on investigating Deep Learning and other machine learning methods to support optical physics applications.

This research project aims to identify and employ state-of-the-art pattern recognition methods to recognize optical and physical properties in image-based experiments. Moreover, make use of artificial intelligence to enhance experimental setups and propose novel experimental techniques.

Deep Learning Applied to Astronomy.

Sat, 27 Apr 2019 00:00:00 +0000

This project investigates the usage of deep learning and machine learning methods to characterize and describe astronomical objects based on images and videos captured by astronomical surveys. This project aims to propose novel methods for data augmentation, recognition methods, and catalog methodologies, exploring supervised and semi-supervised learning methodologies.

2.5D Cartoon Model.

Wed, 27 Apr 2016 00:00:00 +0000

Cartoon 2.5D model is an innovative approach to vectorized 2D illustrations that can rotate in the 3D space. This type of model holds the advantages of 2D vectorized art (capable of translating and scaling without a perceptible loss in quality) while creating animations and interactive poses with low energy expend on monotonous artist work.

The design of a 2.5D model consists of the creation of a few key-frames by the artist. The method automatically creates the inbetweening-frames that allow the model to freely rotate in the 3D space.

This research project investigates methods to improve the 2.5D by providing additional 3D cues that allow different lighting and shading effects, adding depth information to the original 2.5D structure. Additionally, we explore the 2.5D original design to include a bone-like structure, allowing elaborate animations without dramatically increasing the computation cost.

eXtended Reality interfaces.

Wed, 27 Apr 2016 00:00:00 +0000

The eXtended Reality spectrum, which includes virtual-reality and mixed-reality, allows users to experiment on diverse interactive applications with distinctive and sometimes unusual varieties of interactions.

This project aims to study and propose new interactions, always focusing on the usability and the user’s immersion in the XR scene.

The research focuses on developing innovative interfaces and interaction methods that allow the users to interact with the environment naturally and consistently.

Lighting Estimation for Mixed Reality.

Wed, 27 Apr 2016 00:00:00 +0000

Consistent rendering of lighting is a substantial part of the eXtended reality universe, particularly in mixed reality, where real objects are combined with virtual objects. The consistent lighting of real and virtual objects increases the user`s immersion and dramatically improves the user`s experience.

The rendering of consistent lighting requires previous knowledge of the environment lighting. This is a challenging task due to the unknown aspects of real-world lighting settings.

This research project aims to estimate real-world lighting using state-of-the-art machine learning techniques. Exploring Deep Learning methods, combining with computer graphics expertise, we aim to extract valuable information about the environment lighting to provide a consistent mixed reality world.