CMU Portugal Program at Ciência 2022

Between May 16th and 18h, the CMU Portugal Program participated at the Encontro Ciência 2022 Summit to promote its latest initiatives to the public and present some of the research projects developed under the Program. Ciência 2022 is a major Science Conference organized by the Portuguese Government through the Ministry of Science Technology and Higher Education (MCTES), FCT and Ciência Viva held annually in Lisbon to highlight the main achievements in science and technology in Portugal.

On May 16th, CMU Portugal hosted the session “Knowledge Creation and Talent Development under the CMU Portugal Program” and, throughout the event, demonstrations from three CMU Portugal projects (Safeforest, iFetch and MAIA). To present and discuss the Program’s Education initiatives and Doctoral Programs of the 3rd Phase, the CMU Portugal Session counted with five Ph.D. students and alumni who shared an enthusiastic overview of their research work and their experience under the dual-degree and affiliates Ph.D. initiatives.

The session started with a brief presentation from CMU Portugal Co-Director Inês Lynce, one of the two panel moderators, who shared an overview of the Program’s Educational activities throughout the years. João Magalhães, session co-moderator and CMU Portugal Faculty member at FCT UNL introduced the first speaker, John Mendonça.

John Mendonça, a CMU Portugal Affiliated Ph.D. candidate in Electrical and Computer Engineering specializing in Natural Language Processing at Técnico, was one of the first 12 students selected under the 1st edition of the Affiliated Ph.D. Programs initiative. John has a MSc in ECE and worked during his master’s at Defined.ai, developing methods to detect crowdsourcing fraud. He is currently working on CMU Portugal Large Scale project MAIA at INESC-ID and led by Unbabel.

In his presentation he briefly introduced his research work that consists of developing a multilingual chatbot using machine translation.

Being an affiliated Ph.D. student, I can spend up to one year at CMU, where I plan to do research with my co-supervisor in the United States, Alan Lavie, and obviously connect with other like-minded Ph.D. students and faculty at the Language Technologies Institute with Dialogue Systems research in mind.
John Mendonça

Maria Casimiro is a Dual Degree Ph.D. candidate in Software Engineering at Técnico/INESC ID and CMU. She received her MSc in Electrical and Computer Engineering in 2018 from Técnico and has been focusing on the problem of understanding the cost/benefits trade-offs of retraining Machine Learning (ML) models, a task now relevant for most companies. Her research is also focused on fraud detection and has been contributing to CMU Portugal Project CAMELOT led by Feedzai. Last year she was invited to do a summer Internship at the company, and she is currently using their use case of fraud detection in her research work. During her presentation, she highlighted how this experience was important to learn the differences of doing research in a company or academia, to see how they organize and plan to adapt their best practices in her research methods.

Double CMU Portugal journey has been extremely fruitful, and I am thrilled to be part of the Program and to have been given this opportunity.
Maria Casimiro

Alex Gaudio is a 5th-year CMU Portugal Ph.D. student in Electrical and Computer Engineering at FEUP and CMU. Alex completed his bachelor’s degree in Music at Bard College but gave up on being a jazz musician to search for an industry that was more in demand, and Software Engineering was at the top of the list. He soon started working as a Data Scientist and Engineer in some NYC startups, where he co-founded the non-profit NYC Makerspace. At the time, not satisfied with where he was professionally, he decided to give himself a 6 months break to think about his future and at some point, someone suggested he do a Ph.D.

That was a turning point in my life in many things. I realized that at some point you must stop riding a wave and to do things for yourself.  Right then and there,
I set the direction to apply for a Ph.D.
Alex Gaudio

Five years after starting his Ph.D., he is still happy with his choice. Alex’s research work is focused on explainable machine learning.  He was previously participating in the CMU Portugal ERI project SCREEN DR focused on ML applied to Diabetic Retinopathy Detection. Currently, Alex integrates the research team of CMU Portugal’s project TAMI that is developing next-generation explainable machine intelligence in medical image analysis.

Susana Brandão, is a CMU Portugal Alumnus in ECE with a background in Physics who started her Ph.D. in 2009. After her Masters she decided to get a job in Space Engineering, where she was well paid but soon got bored, so she decided to go back to her previous dream of getting a Ph.D. At that time, she met a colleague working at Instituto Superior Técnico with CMU Portugal faculty member João Paulo Costeira and suggested that she would talk to him. He introduced her to the CMU Portugal Program, and she loved the opportunity of studying abroad while maintaining a connection to Portugal.

It was very important for me this opportunity to go abroad and have a degree by Carnegie Mellon but also to be in Portugal to finish my Ph.D. That meant that I would be able to create a network in Portugal and this is very important.People probably underestimate the power of networking, but I think this is really valuable. This is what CMU Portugal offered me. To go abroad and build a network there but, since I wanted to stay in Portugal, I also had the opportunity to have it here.
Susana Brandão

She defended her Ph.D. Thesis on Multiple 3D View Object Recognition, in June 2015, supervised by Manuela Veloso at CMU and João Paulo Costeira in Portugal. After finishing her Ph.D., she worked at Siscog in optimization and afterward at NOS as a Lead Data Scientist. According to her, “lots of doors open when you have a Ph.D. from Carnegie Mellon” and she recently transitioned to Feedzai as an Expert Data Scientist. She is also an Invited Professor at Nova School of Business and Economics and Católica.

Zita Marinho was the last speaker of the day. She is a CMU Portugal program Alumnus in Computer Science/Robotics who graduated in 2018. Her research areas are machine learning and optimization with applications in Robotics and Natural Language Processing. Her thesis was centered on spectral methods for prediction and planning tasks. According to her, one of the best things about CMU is the people you meet. One of the best pieces of advice she got at the time was “talk to everyone. Everyone around you is super interesting and are doing great things. In the future, you might be able to collaborate, so just be friends with them”.

 Your Ph.D. is a marathon, don’t tire yourself out in the first minutes. (…)  it will be hard but if you manage to finish you will gain a lot of things, you gain the knowledge and the tools you learn along the way but also the resilience you will need in the future. 
Zita Marinho

When she finished her Ph.D., she was looking for a regular job but soon after she started, she wanted to go back to research. “It’s ok to fail and make the wrong decisions; you can always correct them”. Zita is currently a Research Scientist at Deepmind and previously worked as Head of Research at Priberam Labs. She is affiliated with the Institute for Robotics and Systems at the University of Lisbon, Portugal.

The session was a great platform to share experiences and disclose what students can expect during their Ph.D. and also after graduation, thanks to the testimonies of the two Ph.D. alumni. At the end of the session there was still time for questions by the audience in a discussion led by Inês Lynce and João Magalhães.

During the event, the CMU Portugal Program held a stand to promote its initiatives to the public and show some demonstrations of its Large-Scale Collaborative projects:

May 16: SAFEFOREST, developing a way to control and reduce large forest fires


The Safeforest project is developing an advanced robotic system for Forest Cleaning and Fire Prevention.  The project intends to create semi-autonomous mobile platforms that can execute a land clearing mission to drastically reduce costs associated with the maintenance of private or public forests, plus control and reduce the propagation of large forest fires.

Promoter: INGENIARIUS
Industrial Co-Promoter: SILVAPOR
Academic Co-promoters: ISR Coimbra and ADAI
CMU: Robotics Institute

 

 

 

May 17: iFetch, creating a new generation of task-oriented conversational agents

The iFetch project proposes to deliver a new generation of task-oriented conversational agents that interact and support users seamlessly using verbal and visual information. This innovative project will allow interaction with consumers through a multimodal interface, providing an experience close to the one you have in a physical store.

Promoter: FARFETCH
Academic Co-promoters: FCT NOVA and Instituto Superior Técnico
CMU: Language Technologies Institute

 

 

May 18: MAIA, enabling chats to develop multilingual conversations
The MAIA project will develop a multilingual conversational platform supported by machine translation and dialogue systems, where AI agents assist human agents. This approach will overcome the limitations of existing customer service.

Promoter: UNBABEL
Academic Co-promoters: Instituto de Telecomunicações and INESC-ID
CMU: Language Technologies Institute

 

Additionally, CMU Portugal Affiliated Ph.D. student Tamás Karácsony at FEUP and INESC TEC submitted a poster now being displayed at Encontro Ciência with the title “3D Motion capture technologies for clinical patient monitoring – a short summary” that was exhibited at Ciência 2019 digital screens.