Our paper titled “Brain-computer interface for generating personally attractive images” has been accepted for publication by IEEE Transactions on Affective Computing. Watch out for a preprint.
Our paper titled “Collaborative Filtering with Preferences Inferred from Brain Signals” has been accepted for publication at The Web Conference (WWW). Watch out for a preprint.
Our paper titled “Spoken Conversational Context Improves Query Auto Completion in Web Search” has been accepted for publication in ACM Transactions on Information Systems. Watch out for a preprint.
Our research was reviewed by the Cognitive Neuroscience Society, see here.
The group starts to operate also at University of Copenhagen, Denmark. We are looking forward for new collaborations and exciting research!
Our paper on connecting generative adversarial networks (GANs) with brain-computer interfaces, titled Neuroadaptive modelling for generating images matching perceptual categories is published by Scientific Reports (Nature). The paper is available here.
Our paper titled “Generating Images Instead of Retrieving Them: Relevance Feedback on Generative Adversarial Networks” published at SIGIR 2020The paper is available here
Our paper on interactive GANs accepted for publication at SIGIR 2020. Stay tuned for a preprint.
Our paper on brain responses and information theory, titled "Information gain modulates brain activity evoked by reading" accepted for publication at Scientific Reports (Nature). The paper is available here.
Our paper on brainsourcing, titled "Brainsourcing: Crowdsourcing Recognition Tasks via Collaborative Brain-Computer Interfacing" accepted for publication at CHI 2020. Paper available here.
We conduct fundamental research into developing new types of implicit brain-computer interfaces and applying them in novel, adaptive and interactive systems. Our current interaction with information retrieval systems rely on explicit interaction. Could we mine the relevance to or interest of the user directly from the human mind? Our research shows, for the first time, that with the help of EEG interpreted via machine learning is indeed possible.
We develop interactive machine learning methods for cognitive modeling of information retrieval, visualization, and information generation, and systems implementing these in real-world information seeking applications. We have developed a technique called interactive intent modeling that allows humans to direct exploratory search: the technique has been implemented in a real-world search engine SciNet.
We develop machine learning models to model, understand, and decode cognitive states of humans interacting with large information spaces, such as texts, images, and video. We have discovered that human information processing of language stimuli when reading follows information theoretic principles and can be explained by information gain.
(now officially with Niklas Ravaja's group)
(supervised with G. Jacucci)
Carlos de la Torre Ortiz
(Interaction laboratory), on leave