Below we present a list of talks with available slides, presented by the Principal Investigator, Marius Leordeanu, on the topic of learning to describe vision in natural language. The material presented constitutes the motivation, scientific basis and some of the results for the Vision in Words project. All the slides are available here. More information about the recent talks, presentations, TV and Radio appearances of the Principal Investigator in science related programs can be found on the PI's webpage here.
Computer Vision, Deep Neural Networks and Robotics: From Learning in Space and Time to Applications, May 25, 2020.
We had the great honor and pleasure to have as guest, together with distinguished professors from the management of the Institute of Mathematics of the Romanian Academy, the Minister of European Funds, Mr. Marcel Bolos, to whom the PI presented our group's research in computer vision, machine learning as well as our research on the connection between computer vision and natural language processing. It was a day of celebration and an important step in the development of artificial intelligence in Romania!
The official government press release of the visit can be accessed here. The official press release of the Romanian Academy of the event can be accessed here. The slides of the PI's presentation can be found here.
Intelligence, beyond the human dimension. Philosophical perspectives on artificial intelligence (Romanian: Inteligența, dincolo de dimensiunea umană. Perspective filosofice asupra inteligenței artificiale), October 2, 2020
We had another great time at the "Izvoare de Philosophie" (Sources of Philosophy) Radio Show at Radio Romania Cultural, together with Cornel Moraru, the host and producer of the program.
We talked about recent advances in artificial intelligence and machine learning and drew parallels between human and computer intelligence. We also had the opportunity to introduce to the audience the PI's recent book on Unsupervised Learning in Space and Time (Springer, 2020). Some of the questions we addressed during the radio program were: How is it possible for humans and machines to learn in an unsupervised fashion? What is context and how does it relate to the meaning of a specific word: object or class of an object? What is the relation between the philosophical, cognitive and computational views on intelligence and consciousness?
Our discussion and the program can be heard in full on the website of this edition, here (in Romanian). The introduction of our radio meeting can be translated as follows:
"Since antiquity, philosophers have tried to understand the way man thinks and learns, by appealing to the usual tools of the human intellect and reflective analysis. That is why, traditionally, we have become accustomed to attributing these complex processes only to human intelligence or, in a broader sense, to some of the creatures with which we share this planet. In recent years, however, with the development of artificial intelligence, more and more such processes are being made better and faster by computer algorithms running in the background of our computers, thus causing a paradigm shift in the way we we relate to humanity. Are artificial intelligence algorithms just an imitation of human intelligence or a completely different type of intelligence? How is it possible for an algorithm to see and learn? Is the unsupervised learning process a conscious one?"
Towards Unsupervised Learning in Space and Time: A Computer Vision Approach speaker at the Eastern European Machine Learning (EEML) summer school and the Summer School on Medical Imaging and Applications (SSIMA), Bucharest, 2019. Slides
Mining for Meaning: From Vision to Language through Multiple Networks Consensus. Invited presentation at the Dagstuhl seminar on Joint Processing of Language and Visual Data, 6-11 January 2019, in Wadern, Germany. Slides
Mining for meaning: the world behind vision, Invited Speaker, international conference on Research Advances in Artificial Intelligence (RAAI) Bucharest, 2018. Slides
ConvNets and Vision, Invited Speaker at the Transylvanian Machine Learning Summer School, Cluj Napoca, July 2018. Slides
The Story behind the Image: When Vision meets Language, Keynote Talks, Botvertising Conference, Romanian Banking Institute, Bucharest, April 2018 and International Summer School on Imaging for Medical Applications, Sibiu, 2018. Slides
A guide to the dreams of tomorrow, Invited speaker at TEDxUniversitateaPolitehnicaBucuresti, March 2018. My talk can be seen on YouTube.
Vision in Words: in Search for the Link Between Vision and Language , Keynote Talk, Bucharest AI Meetup, March 2018. Slides