Monday, 13. June (Pre-C. Day 1)
Workshop on the representation, sharing and evaluation of multimodal agent interaction
-
nteraction is a real world event that takes place in time and physical or virtual space. By definition, it only exists when it happens. This makes it difficult to observe and study interactions, to share interaction data, to replicate or reproduce them and to evaluate agent behaviour in an objective way. Interactions are also extremely complex, covering many variables whose values change from case to case. The physical circumstances are different, the participants are different, and past experiences have an impact on the actual event. Besides, the eye(s) of the camera(s) and/or experimenters are another factor with impact and the man-power needed to capture such data is high. Finally, privacy issues make it difficult to simply record and publish interaction data freely.
It is therefore not a surprise that interaction research progresses slowly. This workshop aims to bring together researchers with different research backgrounds to explore how interaction research can become more standardised and scalable. The goal of this workshop is to explore how researchers and developers can share experiments and data in which multimodal agent interaction plays a role and how these interactions can be compared and evaluated. Especially within real-world physical contexts, modelling and representing situations and contexts for effective interactions is a challenge. We therefore invite researchers and developers to share with us how and why you record multimodal interactions, whether your data can be shared or combined with other data, how systems can be trained and tested and how interaction can be replicated. Machine learning communities like vision and NLP have made a lot of fast progress by creating competitive leaderboards based on benchmark datasets. But although this is great for training unimodal perception models, obviously such datasets are not sufficient for research involving interaction where multiple modalities should be considered.
Heterodox Methods for Interpretable and Efficient Artificial Intelligence
-
In this workshop we will discuss (machine learning) architectures where the human involvement in the design of the model and its data ingestion process allows for both more energy efficient and more interpretable outcomes. Examples of such systems stretch from pure grammatical inference methods and probabilistic programming, where the model (family) is entirely constructed by human hands and only very specific model parameters are learned from data, to various types of interpretable neural network approaches where the specific workings of the output system is much less defined a priori. The goal is to spread knowledge about lesser known approaches to learning from data that use an increased level of human involvement, require less training data, and are tailored to achieve interpretable results in a more efficient way.
Imagining the AI Landscape after the AI Act
-
In April 2021, the EU Parliament published a proposal, the AI Act (AIA), for regulating the use of AI systems and services in the Union market. If adopted, the AIA will have a significant impact in the EU and beyond.
This workshop aims at analyzing how this new regulation will shape the AI technologies of the future. Do we already have the technology to comply with the proposed regulation? How to operationalize the privacy, fairness, and explainability requirements of the AIA? To what extent does the AIA protect individual rights? How is it possible to deliver a process that effectively certificates AI?
This workshop will bring together researchers and practitioners from academia, industry and anyone else with an interest in law and technology to exchange ideas on the multi-faced effects of the AI Act proposal. Paper submissions with an interdisciplinary orientation are particularly welcome, e.g., works at the boundary between AI, human-computer interaction, law, and ethics.
Common Ground Theory and Method Development workshop
-
Exploring, Understanding, and Enhancing Human-Centricity in Hybrid Work Settings
Numerous disciplines contribute to hybrid intelligence work environments, leading to different basic understandings of what exactly human-centered AI means. These understandings are not necessarily rooted in explicit theories, but result from theories in use that lead to a set of methods and instruments that are applied in R&D projects and transferred to practice. The aim of the workshop is to identify a common ground for human-centricity in hybrid work settings from the perspective of different disciplines and research communities involved in specific job design with hybrid intelligence. Therefore the workshop invites (1) theoretical outlines of human-centered hybrid-intelligent work settings, (2) methods, instruments, and standards as theories in use, (3) use cases describing human-centered AI in the workplace. The workshop will conclude with reflections on a special joint issue to discuss the Common Ground Theory. Submissions from tandems of researchers and practitioners are highly appreciated in this third line.
Lunch
-