The SpaceTimeVision research team from University of Bucharest, October, 2021, Bucharest.
The research team's first physical "pandemic" gathering respecting the legal safety measures.
Beautiful landscapes of our PRECIS campus at the University Politehnica of Bucharest, alongside members from our research team captured during one of our flights on a warm autumn day in October.
Our SpaceTimeVision research team in action at the PRECIS research center, Bucharest.
December 18, 2019: Iulia Duta and Andrei Liviu Nicolicioiu presenting our paper, which introduces the Recurrent Space-time Graph Neural Networks model, at Neural Information Processing Systems (NeurIPS) conference, in Vancouver, Canada. (Image on the left)
Students of the world studying Machine Learning at EEML 2019, in Romania
July 5th, 2019: This is one of the most beautiful pictures I have taken in many years! Seeing all our wonderful students and guests from the Eastern European Machine Learning Summer School 2019, here in our beautiful park Herastrau (Bucharest, Romania), is absolutely fantastic!
Our fun and fruitful collaboration with Petru Cercel high school in Targoviste, Romania. We thank all 16 hardworking volunteers that participated in labeling the largest aerial video dataset for semantic segmentation, Ruralscapes.
More details can be found in our paper:
A. Marcu, V. Licaret, D. Costea and M. Leordeanu, Semantics through Time: Semi-supervised Segmentation of Aerial Videos with Iterative Label Propagation, ACCV 2020
At Dagstuhl Seminar, in January 2019, where I presented our work on learning to describe videos in natural language using multiple networks consensus.
More details can be found in our paper:
I. Duta, A. Nicolicioiu, V. Bogolin and M. Leordeanu, Mining for Meaning: from Vision to Language through Multiple Networks Consensus, BMVC 2018
Our team having a great time while learning to fly the drone for the very first time.
Our team got Best Presentation Award for our ICCV article, Creating Roadmaps in Aerial Images with Generative Adversarial Networks and Smoothing-Based Optimization.