Perttu Hämäläinen’s research while affiliated with Aalto University and other places

... Virtual and augmented reality are increasingly used in various domains for guiding body motion, such as improving dance performance [17,1], cycling [14], basketball free throws [18], hand gestures [9], yoga [12], and physical rehabilitation [6,10,2]. These systems situated visualizations with respect to the person or a digital copy of the person to provide intuitive, spatially-relevant feedback. ...

Reference:

Understanding User Needs for Injury Recovery with Augmented Reality

... For example, as individuals age or experience changes in life circumstances, maintaining daily steps goals can become more challenging. However, staying active in different ways other than only taking steps has significant benefits for both physical and mental well-being [7,20]. This shifts the focus from using technologies to support individuals in managing their health to designing technologies to support better those who are already experiencing changes in their lives [10]. ...

Reference:

Life in Transitions: The Role of Technology in Supporting Well-being in the Heart of Change

... Noshaba Cheema and etc explore newly developed technologies in virtual character animation and motion synthesis, putting an emphasis on deep reinforcement learning for generating realistic fatigue animations with not need to capture explicit data. They also discuss a physics-based full-body approach, making it possible to generate efficient real-time interactive animation for virtual characters of varied physiologies [13]. ...

Reference:

Character Motion Synthesis: A Survey

... A prominent research line has focused on creating taxonomies to categorize locomotion techniques [1], including computational algorithms for clustering [33]. Other studies experimentally evaluate novel locomotion implementations against existing methods [6,28]. Since VR is predominantly experienced in first-person view, most research also concentrates on locomotion techniques from an egocentric perspective. ...

Reference:

Effects of Third-Person Locomotion Techniques on Sense of Embodiment in Virtual Reality

... Further, it is noted that even though GAI could be used to address inequity and social asymmetry, it also has the potential of weding the gap of equity (Currie, Hawk & Rohren, 2024). The theme of Well-being is supported by codes from Currie, Hawk and Rohren (2024), Reddy (2024), Vimpari et al. (2023), andLin (2022). The codes highlight potential positive impacts of GAI on social sustainability. ...

Reference:

Generative Artificial Intelligence and the Impact on Sustainability

... Are those requirements fulfilled for that specific dataset? For example, privacy-focused synthetic data must ensure confidentiality and maintain the meaningful statistical properties of the original dataset (Arnold and Neunhoeffer, 2021), while synthetic data used to evaluate research questionnaires must provide plausible answers and be indistinguishable from those of human participants (Hämäläinen et al., 2023). The use of synthetic data is indeed a multidimensional issue; where ethics, performance, and privacy are just some of the concerns when using artificially generated data (Kowalczyk et al., 2023). ...

Reference:

Integrating Generative AI into Information Systems Research: A Framework for Synthetic Data Evaluation

... Parallel to technological exploration, there is an imperative need to develop robust ethical frameworks for AI utilization in architectural education. This aspect of research, drawing inspiration from the perspectives provided by Vimpari et al. (2023), will delve into formulating guidelines that ensure responsible and transparent AI practices. Such guidelines will be crucial in maintaining the integrity of design processes while harnessing AI's capabilities. ...

Reference:

Integrative Analysis of Text-to-Image AI Systems in Architectural Design Education: Pedagogical Innovations and Creative Design Implications

... Therefore, these models use Reinforcement Learning (RL) methods where a control policy is learned automatically overcoming the need for large amounts of labeled training data [61]. Such approaches have been used to model various tasks [61], such as pointing [33], menu search [9] or touchscreen typing [40]. ...

Reference:

Computational Models for In-Vehicle User Interface Design: A Systematic Literature Review

... Therefore, it is necessary and meaningful to predict game difficulty accurately. Most current studies utilize success rates, life points, and completion times to assess game difficulty [10][11][12]. Although they can predict game difficulty to some extent, they hardly take user data into account. ...

Reference:

Game Difficulty Prediction Based on Facial Cues and Game Performance

... Similarly, when we interviewed professionals, several indicated to start from simplified 'non-technology' experiences. For instance, rather than smart sensors, using a marble in a transparent tube to simulate the use of an accelerometer [13]. In workshops with physiotherapists, a too strong focus on technology was often detrimental to a focus on sensations and correction strategies. ...

Reference:

Technology, Movement, and Play Is Hampering and Boosting Interactive Play