Moacir A. Ponti

I explore fair, explainable, and human-centered ways to use machine learning to make technology truly serve people. At Mercado Livre, I lead strategy and research on foundation models, AI agents, and scalable machine learning systems for production. As an Associate Professor at the University of São Paulo (ICMC/USP), my work focuses on deep and representation learning across multiple data modalities. I am a recipient of the Google Latin America Research Award and a CNPq Productivity Fellow since 2018, and have been listed among the top 2% most influential scientists worldwide (Stanford–Elsevier C-score).

Main Research fields: Machine Learning/Representation Learning;  Multimodal domains: audio, speech, image and tabular data.

Recent/highlighted papers (see Publication for more details):

• YourTTS: Towards zero-shot multi-speaker TTS and zero-shot voice conversion for everyone (ICML 2022)

• Sketchformer: Transformer-based representation for sketched structure. (CVPR 2020)

• Everything you wanted to know about Deep Learning for Computer Vision but were afraid to ask (SIBGRAPI Tutorials 2017: papers, slides, code)

Check out my Book on Machine Learning (Theory and Practice), details below!

Previous experiences/education:
Academic visitor (2016-2017); Centre for Vision, Speech and Signal Processing (CVSSP), University of Surrey, UK.
Assistant Professor (2009-2010); Universidade Federal de Viçosa, Rio Paranaíba, MG, Brazil.

PhD (2008) and MSc (2004) at the Universidade Federal de São Carlos (UFSCar), SP, Brazil.
PhD internship (2007) at IEETA, Universidade de Aveiro, Portugal.