David Semedo - Ph.D.
Assistant Professor, School of Science and Technology - NOVA University of Lisbon

My work focuses on bridging vision and language for multimedia understanding. I research deep learning approaches to address the semantic gap between visual and textual modalities, by leveraging on media context. Currently working on Deep learning models for Multimedia and Multimodal Dialog systems.

Expertise and Interests: Multimedia Understanding. Multimodal machine learning. Neural-based Representation Learning. Computer vision and natural language processing. My core research interests lye around the field of multimedia understanding, namely in the intersection of computer vision and NLP, Deep Neural Networks, Artificial Intelligence and Data Mining.

Current positions (M.Sc. and Ph.D.): I’m currently looking for motivated students. If you are interested in working on the fields of deep learning for multimedia, vision and language understanding (dialog systems, visual question-answering, image-captioning, etc.) or neural representation learning, drop me an email!


  • Multimedia Understanding
  • Deep Learning for Multimedia
  • Multimodal Machine Learning
  • Data Mining


School of Science and Technology - NOVA University of Lisbon
2016 - 2020
Ph.D. Deep Learning for Multimedia
School of Science and Technology - NOVA University of Lisbon
2013 - 2015
M.Sc. Metaheuristics for Large-Scale Combinatorial Search
School of Science and Technology - NOVA University of Lisbon
2010 - 2013
B.Sc. Computer Science and Engineering


Recent news and highlights.

Publication Highlights

For a full list of publications see my Google Scholar profile.
Assisting News Media Editors with Cohesive Visual Storylines, 2021, ACM Multimedia
G. Marcelino, *David Semedo*, A. Mourão, S. Blasi, M. Mrak, and J. Magalhães
Open-Domain Conversational Search Assistant with Transformers, 2021, ECIR 2021
Rafael Ferreira , Mariana Leite, David Semedo, João Magalhães
Adaptive Temporal Triplet-loss for Cross-modal Embedding Learning, 2020, ACM Multimedia
David Semedo , João Magalhães
Cross-Modal Subspace Learning with Scheduled Adaptive Margin Constraints, 2019, ACM Multimedia
David Semedo , João Magalhães
Diachronic Cross-modal Embeddings, 2019, ACM Multimedia
David Semedo , João Magalhães
Temporal Cross-Media Retrieval with Soft-Smoothing, 2018, ACM Multimedia
David Semedo , João Magalhães


Brief list of projects in which I have been/I am currently involved.
NewsVisualSeek - Interactive Information Seeking Environments based on Deep State-aware Multimodal Embeddings, NOVA LINCS, 2021-2022
TWIZ - Amazon Alexa Prize 2021 Taskbot, 2021-2022
iFetch - MultiModal Conversational Agents for the online fashion marketplace, CMU-Portugal, 2020-2023
GoLocal - From monitoring global data streams to context-aware recommendations, CMU Portugal, 2019-2020
COGNITUS - Converging broadcast and user generated content for interactive ultra-high definition services, H2020, 2016 - 2018