I am a software developer with a focus on Artificial Intelligence, Computer Vision, and Natural Language Processing (NLP). Throughout my professional journey, I have built and deployed solutions for startups and large companies, utilizing AI, Computer Vision, and NLP technologies. I hold a degree in electrical engineering and a Master's in Computational Intelligence. In addition to product development, I have also successfully conducted research in 3D pose estimation and context recognition using data from wearables.

Work
Experiences

• April 2023 - Present
• Sidia
• Data Scientist
• I’m actively engaged in crafting solutions for Samsung’s Industry 4.0 factory using Computer Vision and AI technologies. My role entails collaborating with the client to define the product’s structure, developing the solution, and overseeing its deployment within the factory. Specifically, I am overseeing a project centered around visual inspection, which is integrated into the production line and aims to enhance quality control.

• October 2022 - April 2023
• FCx Labs
• Data Scientist
• I developed workflows for ETL process. Architecture and implementation for personal-based recommendation systems. Improvements in the search mechanism. APIs for Recommendation systems. EDA for user interactions. Indexing and search of texts and vectors using Elasticsearch. Evaluation Methods for recommendation systems.

• May 2021 - October 2022
• Voxar Labs
• Machine Learning Developer
• SABIA (Centro de Excelência em Saúde & Bem-Estar e Inteligência Artificial) project in partnership with Samsung Inc. We build AI solutions for health using context adaptation of data obtained from wearables devices;
• HP ScalingML project in partnership with HP Inc. We analyze and propose methodologies and tools related to the CI/CD process with a focus on Machine Learning. Furthermore, applying the best methods in the Machine Learning lifecycle. This approach is also known as MLOps.
• HP Collab project in partnership with HP Inc. I worked with 3D face reconstruction, facial keypoints detection, and face reenactment.

• July 2019 - April 2021
• Senfio
• Computer Vision Engineer
• June 2020 - April 2021
• AI Research, Development, and Deployment, related to a face recognition system. Face encoding uses a model based on facenet (face recognition system developed by Google). The system runs in the google cloud platform’s VM through a docker container. Images taken from an ESP32 are sent to this application on the cloud.
• -
• Computer Vision intern
• July 2019 - May 2020
• Selected to be part of Team Developer, building an automatic system for insect counting using Computer Vision and Machine Learning.

Education

• 2021-2023
• Centro de Informática - Federal University of Pernambuco (UFPE)
• MSc in Computational Intelligence, GPA 10/10
• Relevant coursework: Machine Learning, Deep Learning, Reinforcement Learning, Computer Vision, Digital Image Processing, Computer Graphics.

• 2015-2019
• Federal University of Pernambuco (UFPE)
• BS Electronic Engineering, 2nd out of 40 students, GPA 9.17/10
• Relevant coursework: Image Processing, Adaptive Filtering, Digital Signal Processing, Digital Systems, Digital Electronics.

Publications

• Pedro de Souza, Diógenes Wallis de França Silva, Isabella de Andrade, Júlia Dias, João Paulo Lima, Veronica Teichrieb, Jonysberg P. Quintino, Fabio Q. B. da Silva, André L. M. Santos: A Study on the Influence of Sensors in Frequency and Time Domains on Context Recognition. Sensors 23(12): 5756 (2023)

• Diógenes Wallis de França Silva, João Paulo Silva do Monte Lima, Diego Gabriel Francis Thomas, Hideaki Uchiyama, Veronica Teichrieb: UMVpose++: Unsupervised Multi-View Multi-Person 3D Pose Estimation Using Ground Point Matching. VISIGRAPP (4: VISAPP) 2023: 607-614

• Diógenes Wallis de França Silva, João Paulo Silva do Monte Lima, David Macêdo, Cleber Zanchettin, Diego Gabriel Francis Thomas, Hideaki Uchiyama, Veronica Teichrieb: Unsupervised Multi-view Multi-person 3D Pose Estimation Using Reprojection Error. ICANN (3) 2022: 482-494

Award &
Achievements

• Moises.ai - 3rd place on Chord Detection Challenge (2021)
• We made a DECIBEL ++ tool, which consists of improvements at DECIBEL from Daphne Odekerken. Actually, our system can detect more chords than only major and minor, we can also detect sevenths and ninths. Beyond that, we make available a google colab notebook that can run the system. The user only needs to load the music file and a JSON is generated with respective chords and their times.

• Awarded by state government funding FACEPE (2018-2019) to do research
• Signal Processing on Graphs: Applications in Sensor Networks.

• 1st Place at Honorable Mention at 2014 Brazilian Public Schools Mathematical Olympiad
• OBMEP is the largest scientific competition in the country. In just over a decade, OBMEP has reached nearly 18 million students and over 47,000 schools, becoming the largest student olympiad in the world.

• Gold Medal at 2013 Brazilian Public Schools Physics Olympiad
• It is a national-level physics olympiad that includes both practical and theoretical exams, covering various topics.

Technical
Skills

• Area of expertise: Deep Learning, Machine Learning, Computer Vision.
• Workflows: Dagster, Airflow.
• Experiment tracking: MLflow, Neptune.
• Frameworks: PyTorch, Tensorflow, Scikit-learn, OpenCV.
• Programming languages: Python, SQL, C, C++, R.
• Virtualization: Docker.
• Cloud Architecture: AWS, Google Cloud Platform.
• Devices: Lattepanda, Raspberry pi3, Raspberry pi0.
• Indexing and search: Elasticsearch.
• CI/CD: gitlab, github actions.
• Languages: Native Portuguese, Advanced English