Diogo Almeida

Learn More
Residual networks (ResNets) have recently achieved state-of-the-art on challenging computer vision tasks. We introduce Resnet in Resnet (RiR): a deep dual-stream architecture that generalizes ResNets and standard CNNs and is easily implemented with no computational overhead. RiR consistently improves performance over ResNets, outperforms architectures with(More)
Recently, convolutional neural networks (CNNs) have been used as a powerful tool to solve many problems of machine learning and computer vision. In this paper, we aim to provide insight on the property of convolutional neural networks, as well as a generic method to improve the performance of many CNN architectures. Specifically , we first examine existing(More)
BACKGROUND Nonlinear optical (NLO) microscopy techniques have potential to improve the early detection of epithelial ovarian cancer. In this study we showed that multimodal NLO microscopies, including two-photon excitation fluorescence (TPEF), second-harmonic generation (SHG), third-harmonic generation (THG) and fluorescence lifetime imaging microscopy(More)
Quantum dots (QDs) are colloidal semiconductor nanocrystals of a few nanometers in diameter, being their size and shape controlled during the synthesis. They are synthesized from atoms of group II-VI or III-V of the periodic table, such as cadmium telluride (CdTe) or cadmium selenium (CdSe) forming nanoparticles with fluorescent characteristics superior to(More)
This paper presents a catalog of smells in the context of interactive applications. These so-called <i>usability smells</i> are indicators of poor design on an application's user interface, with the potential to hinder not only its usability but also its maintenance and evolution. To eliminate such usability smells we discuss a set of program/usability(More)
Each human genome is a 3 billion base pair set of encoding instructions. Decoding the genome using deep learning fundamentally differs from most tasks, as we do not know the full structure of the data and therefore cannot design architectures to suit it. As such, architectures that fit the structure of genomics should be learned not prescribed. Here, we(More)
  • 1