Carlos Pereira
“During the month of October, Carlos Pereira made significant advances in the implementation of a software library for semantic interoperability in energy communities, contributing to the development of a fully interoperable ecosystem of digital services for electricity consumers. This shows high potential for application in several areas and improves on the concepts developed in previous projects (e.g. the European project InterConnect). Carlos Pereira proved to be self-taught, with a strong self-critical spirit, promoting relevant technical discussions within the institution and always contributing to a positive and relaxed work environment.”
– CPES coordinators
You were involved in the implementation of a software library for semantic interoperability in energy communities; could you tell us a little more about this work, and what are the main results?
Semantic interoperability is a common requirement in current projects. If the definition itself is not obvious, the implementation is even more complex – mainly due to the large number of standards that we must comply with. While some of these models date back to the dawn of the Internet, semantic interoperability is still at an early stage in terms of implementation.
The work developed aims to simplify current methodologies, through a new semantic specification, which we call SEMAPTIC, much simpler and more intuitive, both for the data producer and for the consumers. The model is currently being applied within the scope of energy communities in the ENPOWER project, although it can be used in any domain that requires semantic interoperability for data exchanges between services.
What are the main challenges associated with your activities?
The main challenge is common to all activities: not enough time. The “supersonic” speed of day-to-day tasks at CPES requires a lot of creativity to find the time to be… creative. Even if everything is chaotic, having the ability to stop and reflect to solve any given task is crucial to improve performance and simplify methodologies.
Which aspects of your job do you enjoy the most?
People. I am fortunate to work alongside amazing people, with infinite patience and who are always around: whether to discuss ideas during a coffee break, or to roll up their sleeves night and day, and get things going.
How do you comment on this nomination?
I am quite surprised and honoured. Thank you very much to the CPES coordinators. In addition to the work developed, I also consider this a recognition of the way I perceive my responsibilities, and the way I interact with the teams. To all those who put up with me and help me every single day (you know who I’m talking about): this award is ours!
Isabel Rio-Torto
“Isabel Rio-Torto played a vital role in creating tutorial videos for the use of CTM’s SLURM computing platform. This work is very relevant because it allows the use of resources by new users without the need for frequent technical support. Her dedication and quality of the work developed as a PhD grant holder at the time of execution clearly justify her nomination.”
– CTM coordinators
You’re in charge of creating tutorials on the SLURM computing platform; could you tell us a little more about this process, as well as about the platform itself and how it contributes to the Centre?
SLURM is the shared computational infrastructure of the CTM, composed of a cluster of GPUs and CPUs, which can be used to perform complex computational tasks (e.g., multimedia processing or computational learning). With this platform, the Centre’s computing resources are managed automatically and centrally, ensuring an efficient use of shared resources, considering the waiting times for tasks and fair access to resources. However, the efficient use of this platform requires a specific type of learning by the users. With the increasing number of said users, using these resources as efficiently as possible is even more important. In addition, each time a new user was added to the platform, despite the existing written documentation, more experienced users of the platform sometimes felt the need to demonstrate, in a more practical way, how they used it. This led to the idea of creating videos/tutorials to make said learning process easier, faster and more intuitive for new users, but also to reduce the time that more experienced users dedicated to teaching the good practices they acquired with their own experience.
What are the main challenges associated with your activities? How do you balance your work at CTM with your PhD?
By keeping up to date with the latest research developed and published at an increasingly faster pace, namely in my field (AI, computer vision and natural language processing). I believe that a PhD is not only about research; it can (and I risk saying, should) involve teaching, participating in the organisation of events, disseminating scientific results, monitoring of undergraduate and/or master’s degree students, etc. In this sense, I perceive my role at CTM as a natural aspect of my PhD path.
Which aspects of your job do you enjoy the most?
The constant challenges and freedom to explore new ideas.
How do you comment on this nomination?
I confess that I was not expecting it; obviously, I am very grateful to the CTM coordinators for this nomination. It’s always great to know that people acknowledge my work!
Mafalda Castro
“Mafalda has demonstrated a truly exceptional performance at INESC TEC, standing out for her remarkable ability to deal with multiple projects simultaneously, with proven quality and dedication. In addition to being involved in several national and one European projects, Mafalda also found the time and motivation to contribute significantly to the development of the INESC TEC App, showing not only her multi-tasking skills, but also great team spirit and commitment to the institution. Her ability to balance different responsibilities, always with a high standard of quality in all her work, it’s truly remarkable.”
– Lino Oliveira, Area Manager
You were involved in the development of INESC TEC APP. Could you tell us a little more about this process, and how you contributed to it? What would you like to highlight about this recently introduced tool?
INESCTEC APP began as an initiative to dematerialise the physical employee card, but it ended up becoming a tool to support the community in several day-to-day tasks, like checking the cafeteria’s menu. I was involved in this solution from the beginning, contributing to the design of the interface, the implementation stage and participating in several decisions about the operation. It was an iterative and thorough process, towards providing an easy experience for the users, but something that was quite rewarding. I would like to point out that this application will be useful not only for those who are already part of INESC TEC, but also to facilitate the integration of new members of our community. For instance, it allows locating different hubs, consult the employees’ manual, or obtain contact information of certain Centres or services.
You joined INESC TEC in 2020; since you’ve been working here for almost five years, is there any moment, project or objective achieved that you wish to highlight? What do you expect from now on, as a member of the INESC TEC community?
I joined INESC TEC in 2020, as part of the curricular internship within the scope of the Computer Science degree. During these years, I also completed the master’s degree in the same field, which was one of my greatest goals. In the future, in addition to keep contributing to several research projects, I truly hope I can join other initiatives that are crucial to the institution!
Which aspects of your job do you enjoy the most?
I’m lucky to work with an excellent team that shares the same dedication. In addition, I like the wide variety of projects I can work on, as it allows me to explore different areas, while improving and overcoming new challenges.
How do you comment on this nomination?
I am very grateful to my supervisor, Lino Oliveira, not only for this nomination, but for all the trust he has shown since I joined the institution. It was an honour to be able to collaborate in this project that can truly make a difference at INESC TEC. I’d also like to thank the other people who were involved, and whose effort was vital to achieve this milestone.
Marta Vranas
“There are several reasons that support Marta’s nomination. I’d like to highlight her involvement in the iLoF application for the EARTO Innovation Award, which we won in the ‘Expected Impact‘ category, and her leadership of the Pocket-Vet application for the Crédito Agrícola ‘Entrepreneurship and Innovation’ award (finalist). With just over a year at INESC TEC, Marta has contributed to the recognition of the work done by INESC TEC and to the resolution of complicated cases involving multiple international partner entities. Marta has also contributed very positively to welcoming new elements of SAL and represented our team on a greater number of occasions.”
– Daniel Vasconcelos, Manager of the Technology Licensing Office
You’re part of two different award applications, with the iLoF ending up winning the “EARTO Innovation Award”, and the Pocket-Vet application reaching a place in the final (besides receiving the “Food and Nutrition Security” award). How did these applications reflect your work, and what kind of motivation do you feel while being involved in these types of processes?
At SAL, we work with the purpose of disseminating and valuing the work developed by INESC TEC researchers. Knowing that our applications reach the final or win awards – particularly those granted by national and international entities and experts in these areas – is an extraordinary recognition of the quality of the work developed and a validation of these innovative solutions. This type of accolade not only validates the strategic importance of the initiatives, but also reflects our dedication to each of the technologies in a unique way. These applications play a crucial role in building bridges between R&D, industry and society, broadening the reach and future impact of technological innovations.
What are the main challenges associated with your activities?
One of the main challenges is to find the ideal balance between all parties involved, to ensure the success of each process. From the collection and selection of important and relevant information that supports the feasibility and applicability of innovative technology, to the negotiation, e.g. of agreements and licenses with external entities. Although it may seem easy and linear at times, each party involved brings their own expectations, priorities and objectives, so finding common ground and promoting said balance is sometimes one of the most complex parts of our work,
Which aspects of your job do you enjoy the most?
As a biochemist, I have been involved in research for many years, which allows me to have a comprehensive outlook of R&D. Currently, what excites me the most about my work is the opportunity to explore what goes beyond the “technical aspects”, to know new languages and the reality of those who integrate innovations – and what they really need. Concerning SAL, what motivates me the most is the dedication required to address each case. We learn a lot from each technology we analyse, from the partner entities, and from the industry agents. This allows us to improve our experiences and develop robust tools to focus on this area of knowledge and technology transfer.
How do you comment on this nomination?
I’m very proud! It is gratifying to know that people acknowledge our work, and to know that we are contributing positively to the success of the team.
Matias Molina
“Matias’ commitment throughout the EMERITUS project was remarkable. He carried out all his tasks with significant accountability and dedication. Last month, the paper ‘More (Enough) Is Better: Towards Few-Shot Illegal Landfill Waste Segmentation’ (Matias Molina et al.) won the PAIS Outstanding Paper Award at the ECAI 2024 – European Conference on AI – one of the major AI conferences in Europe and the world.”
– LIIAD coordinators
You’re involved in the EMERITUS project; can you tell us a little bit more about this initiative (main objective, differentiating factors, main results, challenges, etc.)?
The EMERITUS project is a large European Union initiative aimed at researching and developing tools to tackle environmental crimes. While the project covers a wide range of activities and involves experts from various fields, our institution focuses on using AI to address different environmental challenges. I am particularly involved in detecting illegal waste dumps through aerial imagery, such as satellite images.
You’re also one of the authors of the paper that received the PAIS Outstanding Paper Award at ECAI 2024; what is the focus of this work, and how does it feel to be involved in an award-winning work, acknowledged during one of the major events in the area?
Our paper focuses on using AI to detect illegal waste dumps through satellite imagery. In the literature, we can find a few AI studies that predict if waste is present in an image, but identifying the exact areas, or “waste segments,” is challenging because AI models generally require vast amounts of labelled data, which is difficult to gather in environmental crime contexts. Our approach combines different AI models and a few annotations to locate waste segments in the image. We emphasised the need for an exhaustive evaluation using different metrics to reflect the quality required in the EMERITUS project. The recognition from ECAI with the PAIS Outstanding Paper Award was a wonderful surprise! It’s not just an honour; I also perceive it as recognition of our efforts in using AI to solve a global issue.
Which aspects of your job do you enjoy the most?
What I like the most about working in science, is that it allows me to continuously satisfy my curiosity. I also enjoy seeing how theories impact real-world, applied problems. It’s also very meaningful to talk with my colleagues and exchange both casual conversations and scientific discussions.
How do you comment on this nomination?
I’m ecstatic with this nomination; it was a pleasant surprise. I feel immense gratitude for all the LIAAD coordinators, especially Prof. João Gama, for his continuous support and trust. I’m also grateful for the opportunity to contribute to meaningful projects. I’d like to share this nomination with my colleagues, who have helped create a positive atmosphere during my time in Porto – an important factor that contributes significantly to our results.
Tamás Karácsony
“The C-BER coordinators would like to nominate Tamás Karácsony – PhD student within the scope of PDEEC/FEUP doctoral programme funded by CMU-Portugal – for winning the ‘IEEE AI Innovation’ award at the GITEX Global 2024, the largest technology and start-ups event in the world, promoted by IEEE Entrepreneurship.”
– C-BER coordinators
You received the “IEEE AI Innovation” award at GITEX Global 2024, thanks to your work in epilepsy. Can you tell us a little bit more about this work (the motivation behind it, the main goal and results achieved, challenges while developing it, etc.)?
Sure! My work is about advancing clinical diagnosis in epilepsy through explainable deep learning. This is an exceptionally challenging computer vision problem because we’re mapping complex, temporospatial 4D seizure movements to specific brain regions in highly data-scarce environments. Clinical practice today still depends on manual visual inspection for semiology-based evaluation, which is subjective, time-intensive, and prone to variability. Our goal is to create a quantitative, automated diagnosis support system using clinical 3D motion capture to make this process faster, more reliable, and more explainable. The research has evolved significantly: we started with an end-to-end action recognition approach that achieved state-of-the-art results, published in Nature Scientific Reports. Recognising the complexities of real-world clinical environments, we addressed challenges like in-bed blanket occlusion by developing synthetic data generation techniques, allowing us to train robust human pose estimation (HPE) architectures. Currently, we’re advancing this work by collecting 4K video data and building a simulated dataset to validate a two-stage action recognition pipeline, which shows great promise and is on track for publication. Collaboration has been central to this project. Working with clinical partners at LMU Munich and CHUSJ, and spending a year at CMU, allowed us to combine expertise in AI, computer vision, and neurology. It’s been an incredible experience to be part of such a diverse, interdisciplinary and international team tackling this complex problem in healthcare.
Of course, challenges remain. Data scarcity is one of the biggest hurdles – clinical samples are inherently limited and difficult to acquire. To address this, we designed the two-stage action recognition approach, which not only leverages transfer learning and synthetic augmentation but also offers several advantages. It efficiently uses scarce clinical data, provides privacy-preserving solutions to enable multi-hospital collaborations, and scales well for larger deployments. Additionally, the system offers movement quantification and explainability, which are critical for building trust and enhancing its utility in clinical settings. At its core, this project demonstrates how cutting-edge computer vision techniques can transform clinical workflows. That’s what excites me most: the opportunity to solve meaningful, real-world problems with technology.
Receiving this award at GITEX was an incredible honour and validation of the work. It’s a testament to the power of interdisciplinary collaboration and innovation. Whether in clinical research or broader applications of AI, this project has the potential to push boundaries in both healthcare and technology.
How does it feel to receive an award at an event like GITEX Global? Must be a major motivation to keep advancing your work, correct? What are your short-term plans concerning the activities you’ve been developing – and that earned you this award?
It feels amazing – being recognised at such a global event really reinforces the importance of this work. It’s a huge motivation to keep moving forward. In the short term, we have multiple publications lined up, we’re analysing intellectual property strategies, and we’re even considering the potential for a spin-off. Of course, I also need to finish my PhD thesis soon since I’m in my last year.
Which aspects of your job do you enjoy the most?
I really enjoy working in this interdisciplinary field of computer vision and neurology. Tackling complex problems and finding solutions that could directly impact patient care is incredibly rewarding. It’s exciting to bridge the gap between technology and medicine.
How do you comment on this nomination?
I’m truly honoured by this recognition, and I’d like to thank to my supervisor, Professor João Paulo Cunha. It reflects the collaborative effort of a multidisciplinary team and the potential impact of our work. This nomination motivates me to continue advancing innovations in AI and healthcare.
News, current topics, curiosities and so much more about INESC TEC and its community!
