David Macêdo

David Macêdo

PhD in Computer Science (Deep Learning)

About

53
Publications
10,006
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
271
Citations
Citations since 2017
50 Research Items
269 Citations
2017201820192020202120222023020406080100120
2017201820192020202120222023020406080100120
2017201820192020202120222023020406080100120
2017201820192020202120222023020406080100120
Introduction
Visiting Researcher with Montreal Institute for Learning Algorithms (MILA), University of Montreal (UdeM), Quebec, Canada. Co-creator and Collaborator Professor of the Deep Learning course of the Computer Science Master and Doctorate Programs at the Center for Informatics (CIn), Federal University of Pernambuco (UFPE), Brazil. PhD Candidate in Computer Science at CIn/UFPE, Brazil. NeurIPS, ICLR, ICML, and IEEE Reviewer. https://dlmacedo.com/
Additional affiliations
September 2019 - August 2020
Université de Montréal
Position
  • Researcher
August 2016 - June 2022
Federal University of Pernambuco
Position
  • Professor
April 2015 - present
Universidade Nova Roma
Position
  • Professor (Full)
Education
March 2018 - June 2022
Federal University of Pernambuco
Field of study
  • Doctor of Philosophy (Computer Science)
March 2016 - July 2017
Federal University of Pernambuco
Field of study
  • Master of Science (Computer Science)
March 1992 - December 1996
Federal University of Pernambuco
Field of study
  • Bachelor of Engineering (Electronic Engineering)

Publications

Publications (53)
Chapter
Full-text available
Web servers provide most internet services, such as information sharing, financial, health, entertainment, and education. In this context, the web has become the principal place for attackers. Unfortunately, most defensive techniques for web servers cannot deal with the complexity and evolution of cyber attacks on HTTP requests. However, machine le...
Chapter
Proper optimization of deep neural networks is an open research question since an optimal procedure to change the learning rate throughout training is still unknown. Manually defining a learning rate schedule involves troublesome, time-consuming try and error procedures to determine hyperparameters such as learning rate decay epochs and learning ra...
Chapter
This work addresses multi-view multi-person 3D pose estimation in synchronized and calibrated camera views. Recent approaches estimate neural network weights in a supervised way; they rely on ground truth annotated datasets to compute the loss function and optimize the weights in the network. However, manually labeling ground truth datasets is labo...
Chapter
Full-text available
The digitalization of society potentialized services provided through the Internet, such as information sharing, entertainment, and education. With the rise of end-user services, we also verify the growth of attacks. Unfortunately, most defensive techniques of Web Intrusion Systems cannot deal with the complexity of cyber attacks on HTTP requests....
Preprint
Building robust deterministic deep neural networks is still a challenge. On the one hand, some approaches improve out-of-distribution detection at the cost of reducing classification accuracy in some situations. On the other hand, some methods simultaneously increase classification accuracy, out-of-distribution detection, and uncertainty estimation...
Article
Context There has been a delay in the exploration of the benefits of deep learning for human action and activity recognition applications. Within these fields, the detection of falls attracts attention due to its excellent public utility. Fall detection can be implemented in facilities such as nursing homes, areas with public cameras, and the homes...
Article
Augmentative and Alternative Communication (AAC) boards are essential tools for people with Complex Communication Needs (e.g., a person with down’s syndrome, autism, or cerebral palsy). These boards allow the construction of messages by arranging pictograms in sequence. In this context, a pictogram is a picture with a label that denotes an action,...
Article
Full-text available
Early alert fire and smoke detection systems are crucial for daily and security management decision-making. Recent literature approaches are based on Deep Learning (DL) models. Efficient models are required for hardware-constrained systems, such as mobile devices, embedded systems, and robotics achieving high performance at low power consumption. F...
Preprint
Full-text available
Expressing and identifying emotions through facial and physical expressions is a significant part of social interaction. Emotion recognition is an essential task in computer vision due to its various applications and mainly for allowing a more natural interaction between humans and machines. The common approaches for emotion recognition focus on an...
Conference Paper
Early alert fire and smoke detection systems are crucial for management decision making as daily and security operations. One of the new approaches to the problem is the use of images to perform the detection. Fire and smoke recognition from visual scenes is a demanding task due to the high variance of color and texture. In recent years, several fi...
Article
Full-text available
In this article, we argue that the unsatisfactory out-of-distribution (OOD) detection performance of neural networks is mainly due to the SoftMax loss anisotropy and propensity to produce low entropy probability distributions in disagreement with the principle of maximum entropy. On the one hand, current OOD detection approaches usually do not dire...
Preprint
Full-text available
Current out-of-distribution detection approaches usually present special requirements (e.g., collecting outlier data and hyperparameter validation) and produce side effects (classification accuracy drop and slow/inefficient inferences). Recently, entropic out-of-distribution detection has been proposed as a seamless approach (i.e., a solution that...
Preprint
Full-text available
Proper optimization of deep neural networks is an open research question since an optimal procedure to change the learning rate throughout training is still unknown. Manually defining a learning rate schedule involves troublesome time-consuming try and error procedures to determine hyperparameters such as learning rate decay epochs and learning rat...
Conference Paper
Full-text available
Most of the automatic fire alarm systems detect the fire presence through sensors like thermal, smoke, or flame. One of the new approaches to the problem is the use of images to perform the detection. The image approach is promising since it does not need specific sensors and can be easily embedded in different devices. However, besides the high pe...
Article
Cyber-attacks on cyber-physical systems (CPSs) can lead to sensing and actuation misbehavior, severe damages to physical objects, and safety risks. Machine learning algorithms have been proposed for hindering cyber-attacks on CPSs, but the absence of labeled data from novel attacks makes their detection quite challenging. In this context, Generativ...
Preprint
Most of the automatic fire alarm systems detect the fire presence through sensors like thermal, smoke, or flame. One of the new approaches to the problem is the use of images to perform the detection. The image approach is promising since it does not need specific sensors and can be easily embedded in different devices. However, besides the high pe...
Preprint
Full-text available
Current out-of-distribution detection (ODD) approaches present severe drawbacks that make impracticable their large scale adoption in real-world applications. In this paper, we propose a novel loss called Hyperparameter-Free IsoMax that overcomes these limitations. We modified the original IsoMax loss to improve ODD performance while maintaining be...
Preprint
Full-text available
Relation extraction (RE) consists in categorizing the relationship between entities in a sentence. A recent paradigm to develop relation extractors is Distant Supervision (DS), which allows the automatic creation of new datasets by taking an alignment between a text corpus and a Knowledge Base (KB). KBs can sometimes also provide additional informa...
Preprint
Full-text available
The Know Your Customer (KYC) and Anti Money Laundering (AML) are worldwide practices to online customer identification based on personal identification documents, similarity and liveness checking, and proof of address. To answer the basic regulation question: are you whom you say you are? The customer needs to upload valid identification documents...
Preprint
Full-text available
Speaker Recognition and Speaker Identification are challenging tasks with essential applications such as automation, authentication, and security. Deep learning approaches like SincNet and AM-SincNet presented great results on these tasks. The promising performance took these models to real-world applications that becoming fundamentally end-user dr...
Preprint
Full-text available
The detection of objects considering a 6DoF pose is common requisite to build virtual and augmented reality applications. It is usually a complex task witch requires real-time processing and high precision results for an adequate user experience. Recently, different deep learning techniques have been proposed to detect objects in 6DoF in RGB images...
Chapter
Full-text available
Embedding artificial intelligence on constrained platforms has become a trend since the growth of embedded systems and mobile devices, experimented in recent years. Although constrained platforms do not have enough processing capabilities to train a sophisticated deep learning model, like convolutional neural networks (CNN), they are already capabl...
Chapter
Full-text available
Document classification is a challenging task with important applications. The deep learning approaches to the problem have gained much attention recently. Despite the progress, the proposed models do not incorporate the knowledge of the document structure in the architecture efficiently and not take into account the contexting importance of words...
Chapter
Full-text available
The recognition of sign language is a challenging task with an important role in society to facilitate the communication of deaf persons. We propose a new approach of Spatial-Temporal Graph Convolutional Network for sign language recognition based on the human skeletal movements. The method uses graphs to capture the dynamics of the signs in two di...
Preprint
Current out-of-distribution detection (ODD) approaches require cumbersome procedures that add undesired side-effects to the solution. In this paper, we argue that the uncertainty in neural networks is mainly due to SoftMax loss anisotropy. Consequently, we propose an isotropic loss (IsoMax) and a decision score (Entropic Score) to significantly imp...
Conference Paper
Os algoritmos normalmente empregados para controle energético em redes IoT envolvem funções de otimização com considerável complexidade e controle rigoroso do ambiente de teste. Isso gera uma lacuna entre o projeto, análise teórica e processamento em tempo real dos dispositivos da rede. O presente artigo propõe uma nova abordagem baseada em aprendi...
Article
Full-text available
A substantial number of expert and intelligent systems rely on deep learning methods to solve problems in areas such as economics, physics, and medicine. Improving the accuracy of the activation functions used by such methods can directly and positively impact the overall performance and quality of the mentioned systems at no cost whatsoever. In th...
Preprint
Full-text available
Speaker Recognition is a challenging task with essential applications such as authentication, automation, and security. The SincNet is a new deep learning based model which has produced promising results to tackle the mentioned task. To train deep learning systems, the loss function is essential to the network performance. The Softmax loss function...
Preprint
The recognition of sign language is a challenging task with an important role in society to facilitate the communication of deaf persons. We propose a new approach of Spatial-Temporal Graph Convolutional Network to sign language recognition based on the human skeletal movements. The method uses graphs to capture the signs dynamics in two dimensions...
Preprint
Full-text available
Most of the research in convolutional neural networks has focused on increasing network depth to improve accuracy, resulting in a massive number of parameters which restricts the trained network to platforms with memory and processing constraints. We propose to modify the structure of the Very Deep Convolutional Neural Networks (VDCNN) model to fit...
Preprint
Full-text available
Cardiovascular diseases are one of the most common causes of death in the world. Prevention, knowledge of previous cases in the family, and early detection is the best strategy to reduce this fact. Different machine learning approaches to automatic diagnostic are being proposed to this task. As in most health problems, the imbalance between example...
Preprint
Speaker Recognition is a challenging task with essential applications such as authentication, automation, and security. The SincNet is a new deep learning based model which has produced promising results to tackle the mentioned task. To train deep learning systems, the loss function is essential to the network performance. The Softmax loss function...
Preprint
Full-text available
Document classification is a challenging task with important applications. Deep learning approaches to the problem have gained much attention. Despite the progress, the proposed models do not incorporate the knowledge of the document structure in the architecture efficiently and not take into account the contexting dependent importance of words and...
Thesis
Full-text available
Recently, deep learning has caused a significant impact on computer vision, speech recognition, and natural language understanding. In spite of the remarkable advances, deep learning recent performance gains have been modest and usually rely on increasing the depth of the models, which often requires more computational resources such as processing...
Conference Paper
Full-text available
The Brazilian System for Advanced Metering (SiBMA) is a solution that defines an application layer protocol designed to meet the needs of the Advanced Metering Infrastructure for the energy market in Brazil. This paper presents the proposed solution from an architectural standpoint. A reference implementation was developed with the goal to validate...
Article
A secret-key cipher based on error-correcting codes is introduced with the distinguishing feature that its error generator is implemented by a non-linear combination of codewords of two linear complementary codes. The error generator produces codewords of a non-linear structured code at its output. The security of the proposed cipher relies on the...
Article
Full-text available
The cryptanalysis of a recently proposed public-key cipher is presented. The mathematical structure of the cipher is based on linear complementary subspaces over a finite field. The cipher is broken simply by multiplying the ciphertext by a matrix which is the multiplicative inverse of a matrix formed with the public information available

Network

Cited By

Projects

Project (1)
Project
Deep Learning Research.