Had a good time at the Lorentz workshop about Collecting, Annotating & Analyzing Video Data, held in Leiden from 30 Oct 2017 through 3 Nov 2017. Interesting discussions about how to improve existing annotation tools for analysis of (social) interactions.
The 7th International Conference on Affective Computing and Intelligent Interaction (ACII2017) was held in San Antonio, TX, 23rd – 26th Oct, 2017. I was social media chair, member of the Senior Program Committee and we had 2 papers accepted. Jaebok Kim had a paper on “Learning spectro-temporal features with 3D CNNs for speech emotion recognition“. Minha Lee presented her poster “Exploring moral conflicts in speech: multidisciplinary analysis of affect and stress“. There was a strong industry panel with Liz Shriberg, Rosalind Picard, Tom Insel, Tanzeem Choudhury, Jennifer Healey, Mary Czerwinski, and Jim Harper. I enjoyed ACII2017 very much and it was great seeing everybody again!
Interspeech 2017 was held this year in Stockholm. It was great seeing everybody again. Jaebok Kim presented our paper on speech emotion recognition “in the wild” using multi-task learning. In our research field, social signal processing, rich big data is key. In general, we see that it is becoming more difficult for academia to compete with multinationals such as Google, Amazon, and Apple in terms of technology development and data acquisition. More funding should become available for academia to collect data that is both high of quality and quantity. I have been discussing with colleagues the impact that the rise of deep learning has had on our language and speech technology research in the past few years – recognition performances have increased remarkably. But this also gives us some food for thought – for example, what are the scientific questions in language and speech (technology) research that we are answering with these new methods, how can we explain the results that we have obtained with these methods?
Our work in collaboration with Jürgen Trouvain (Saarland University) has been mentioned in an article by Scientific American, see https://www.scientificamerican.com/article/scientists-are-teaching-robots-to-laugh/
Interview about being a popstar in science (in Dutch), see https://www.utoday.nl/science/64311/popster-voor-de-wetenschap
Thank you for inviting Cristina Zaga and me for giving talks at the Mintlab in Leuven, see https://www.facebook.com/permalink.php?story_fbid=1497790863588548&id=453794087988236
The 4TU Symposium on Empathic Technologies with keynote speakers Hatice Gunes and Marc Hassenzahl was a great succes! See https://www.utoday.nl/news/64135/emphatic-technologies-that-can-feel-you and https://empathiclighting3tu.wordpress.com/pictures/
I gave talks and interviews at and for Pakhuis de Zwijger, Nemo Kennislink, NPO Radio 1!