cv

Attaché temporaire d'enseignement et de recherche (ATER)

Université Gustave Eiffel

• Teaching

Data Scientist

NeuralMind

• Deep Learning

PhD

ESIEE Paris (Université Gustave Eiffel)

Explainable Artificial Intelligence

MSc

Universidade Estadual de Campinas (Unicamp)

Machine Learning

Bachelor

Universidade Estadual Paulista Júlio de Mesquita Filho (UNESP)

Computer Science

Université Gustave Eiffel

Teaching assistant

• Image Processing (Master)
• Algorithms and Programming (Bachelor)
• C Programming (Bachelor)
• Databases (Bachelor)

EPITA - École d'Ingénieurs en Informatique

Teaching assistant

• Python for Big Data (Master)
• Introduction to Neural Networks (Master)
• Mathematics of the signal (Master)
• Rational Languages Theory (Bachelor)
• Algorithms Complexity (Bachelor)

Unicamp

Teaching assistant

• Complex data mining regarding information retrieval learning (Specialization)
• Complex data mining regarding supervised learning (Specialization)
• Complex data mining regarding unsupervised learning (Specialization)
• Algorithms and computer programming (Bachelor)

Best presentation

Journée des doctorants MSTIC

Academic Merit

Faculdade de Ciência e Tecnologia – UNESP

Bridging Human Concepts and Computer Vision for Explainable Face Verification

Workshop at AI*IA

In this paper, we present an approach to combine computer and human vision to increase the explanation's interpretability of a face verification algorithm. In particular, we are inspired by the human perceptual process to understand how machines perceive face's human-semantic areas during face comparison tasks.

Transforming gradient-based techniques into interpretable methods

Pattern Recogniton Letters - Elsevier

We introduce GAD (Gradient Artificial Distancing) as a supportive framework for gradient-based explainable techniques. Its primary objective is to accentuate influential regions by establishing distinctions between classes. The essence of GAD is to limit the scope of analysis during visualization and, consequently reduce image noise.

Unsupervised discovery of Interpretable Visual Concepts

Information Sciences - Elsevier

In this paper, we propose two methods, Maximum Activation Groups Extraction (MAGE) and Multiscale Interpretable Visualization (Ms-IV), to explain the model's decision, enhancing global interpretability. MAGE finds, for a given CNN, combinations of features which, globally, form a semantic meaning, that we call concepts.

Reasoning with trees: interpreting CNNs using hierarchies

Arxiv

In this paper, we propose a framework to construct model-based hierarchical segmentations that maintain the model's reasoning fidelity and allows both human-centric and model-centric segmentation. This approach offers multiscale explanations, aiding bias identification and enhancing understanding of neural network decision-making. Experiments show that our framework, xAiTrees, delivers highly interpretable and faithful model explanations, not only surpassing traditional xAI methods but shedding new light on a novel approach to enhancing xAI interpretability.

Manifold Learning for Real-World Event Understanding

IEEE Transactions on Information Forensics and Security

We extend upon our prior work and present a learning-from-data method for dynamically learning the contribution of different components for a more effective event representation. The method relies upon just a few training samples (few-shot learning), which can be easily provided by an investigator.

Forensic Event Analysis: From Seemingly Unrelated Data to Understanding

IEEE Security & Privacy

We discuss the problem of restructuring visual data from different heterogeneous sources to analyze an event of interest. We present X-coherence: a pipeline seeking to organize and represent pieces of data, tying them coherently with the real world and with one another. We also outline research challenges while seeking X-coherence.

Image Semantic Representation for Event Understanding

IEEE International Workshop on Information Forensics and Security

We propose an image semantic representation method that helps to understand the discrimination of Representative Images (RI) from Non-representative Images (NRI). Our method, called Event Semantic Space (ESS), generates a low-dimensional image representation by exploiting the semantics of some images with high representativeness and some representative components of the events (e.g., places, objects, and people).