Non-verbal cues are essential in communication, and eye contact and gaze awareness are crucial components of these cues. However, video-based communication lacks many of these non-verbal cues. To address this issue, we have developed a low-cost prototype allowing natural eye contact during one-to-one video calls. In a study focused on telepresence in remote psychotherapy, we demonstrate that our prototype enhances the perception of genuine and realistic communication with higher physical and social presence compared to classical video meeting setups with the camera placed on the screen.

Forced migration affects millions worldwide due to conflict, disasters, and persecution. This paper presents a participatory approach to serious game development, engaging individuals with lived migration experiences to create authentic and impactful narratives. Based on qualitative research, we propose a game design methodology to enhance empathy and raise awareness of forced migration. Our study demonstrates the potential of participatory games to bridge the gap between forced migrants and the broader public, fostering empathy, social change, and empowerment for marginalized communities.

Non-verbal cues are essential in communication, and eye contact and gaze awareness are crucial components of these cues. However, video-based communication lacks many of these non-verbal cues. To address this issue, we have developed a low-cost prototype allowing natural eye contact during one-to-one video calls. In a study focused on telepresence in remote psychotherapy, we demonstrate that our prototype enhances the perception of genuine and realistic communication with higher physical and social presence compared to classical video meeting setups with the camera placed on the screen.

Game jams are events where development teams collaborate to create games in days or weeks, sharing common themes and limitations. However, it is unclear how participants in game jams fit into the larger gaming culture regarding their motivations and identities as gamers. Our paper surveys two game jam populations to determine their player motivation profiles using the Hexad-12 questionnaire, roles in the development teams at a game jam, their correspondence to the gamer and LGBTQ+ identity, and their play-time per week. The results show that the population in our samples does not differ significantly from the one reported in the original Hexad study. We found correlations between the Hexad player types and the development roles at a game jam. People interested in Game/Level Design and people reporting having fun as a reason for attending the game jam were more likely to be Achiever types, people interested in Art were more likely, while those interested in Programming were less likely to be Disruptor types, and the more participants identified as a gamer, the more likely they were Player types, and/or preferred the Game/Level Design role. Furthermore, we found that both samples identified to some extent with the term gamer. Additionally, the more participants leaned towards identifying as a gamer, the more likely they were Players and/or preferred the Game/Level Design hat.

With OpenAI's Dall-E, Midjourney, and Adobe Firefly, computer-generated visual content has hit the mass market. Machine learning-based algorithms can now create, and re-mix multimedia content based on huge corpora of images and videos and relieve creative professionals of tedious work. While this has gained much momentum lately, procedurally generated content (PCG) has been around for quite some time already. Especially in video game development, randomized levels, behavior, aesthetics, and even narratives increase replayability and engage the audience longer. Prominent examples are Minecraft and Diablo, where the game world is randomly generated, and Borderlands, where in-game items are generated on the fly. PCG is applied on multiple levels with different purposes, like generating terrain, weather, road and transport networks, house layouts, puzzles, textures, and mazes, just to name a few. Hedrikx et al. [1] give a comprehensive overview of the topic. In a typical scenario, a mix of algorithms is employed to create content on the fly. Generative grammar algorithms are often employed for vegetation, fractal noise is used to generate terrain and clouds, and simulation is used to create road networks or to erode terrain further. Lately, deep learning-based approaches have become available. The most notable example is AI Dungeon [2], where users converse with GPT in a text adventure. However, large neural networks require significant computational power and are hard to explain and constrain, so procedural content generation tends to use less complex algorithms, where game designers can give hard constraints to influence the outcome.

Writing source code for programs with lightweight text editors or fully featured integrated development environments is considered the main method of programming. Notebooks, however, are an extremely practical tool. In contrast to IDEs, projects are set up more easily and they allow for running programs in a read-eval-print loop (REPL) environment. The Jupyter Notebooks Quick Start Guide [1] describes notebook documents as "… both human-readable documents containing the analysis description and the results (figures, tables, etc..) as well as executable documents which can be run to perform data analysis." Basically, markdown text can be mixed with program source code in a sequence of sections, each dedicated to either programming or description and documentation. Source code sections can be executed and the output is appended to the section, even formatted in the form of graphs, diagrams, or tables. REPL and notebook based environments have proven to be useful in many scenarios including when exploring new libraries and frameworks, to prototype code, or as an educational tool to create interactive lecture material. A prominent example is Jupyter, which is the highly successful project behind Jupyter Notebooks and the recent JupyterLab, i.e. web-based systems to run and share notebooks that contain code, equations, data visualizations, and data exploration and narrative text.

Open-source software is a relevant topic in video game development. Taking a look at the most frequently employed game engines for developing Android games [1] we can see that seven out of ten ranked engines are OSS. Over the last decade, more and more game studios and individual developers switched to open-source software. Oliver Franzke from Double Fine Productions described the phenomenon from his point of view at GDC Europe 2013 [2]. For him developing game engines from scratch is too costly, and re-using self-made game engines often involves too much repurposing as they were developed with one specific game in mind. Existing third-party game engines might not support all features needed, and requesting new features or APIs from the developers is a tedious effort. In contrast to that, an open-source game engine is typically stable, extendable, and - in the best case - has a lively community accepting patches and helping out problems. Best of all, if a bug in the game engine is found or a feature is needed, one can fix it and compile it instead of filing a bug report or feature request and hoping for the next release of the engine being in time for the release of the own project. Epic Games was one of the first AAA game engine developers to publish the source code of their Unreal Engine [3] in 2015 to let game developers investigate critical parts and send in patches and file bug reports. However, they did not change their model to OSS. Amazon (with the Lumberyard engine), Unity, and Crytek (with the CryEngine) follow similar models, where the source code is available for developers, but no open-source license is used. Other prominent source code releases included the famous id Tech engines, which powered games like Doom and Quake, and the Serious Sam Engine from Croteam [4, 5]. In both cases, they were made available GPL license for historical and educational purposes and developers never released the most recent version of their engines. These engines are not used only to develop games, but also to implement simulators that can be used to create datasets used in research, like CARLA [19] (a simulator for autonomous driving research) or UnrealCV [20], that are both based on Unreal Engine.

Performance optimisation continues to be a relevant topic both in hardware and software development, with video games producing fully rendered images every 16 or 34 ms, depending on the desired framerate. Human observers close their eyes for about 300 ms an average of twelve times per minute, which means many frames will never be observed. This paper aimed to examine if it would be possible to reduce rendering time by detecting and skipping these unobserved frames. Blinks were identified during runtime by detecting the eye aspect ratio of the observer in low-quality web camera footage. A prototype using this method was tested on a small group of subjects to determine if footage watched this way was perceived as distracting or of lesser quality than unaltered images. Results from a questionnaire suggest that the altered footage did not impact the subjects’ opinions, with no participant reporting any visual disturbances. Because this test used video footage, skipping frames was substituted by a lower resolution render. Altered frames were rendered an average of five percent faster than their unaltered counterparts.

Artificial intelligence (AI) is predicted to have profound effects on the future of video capsule endoscopy (VCE) technology. The potential lies in improving anomaly detection while reducing manual labour. Existing work demonstrates the promising benefits of AI-based computer-assisted diagnosis systems for VCE. They also show great potential for improvements to achieve even better results. Also, medical data is often sparse and unavailable to the research community, and qualified medical personnel rarely have time for the tedious labelling work. We present Kvasir-Capsule , a large VCE dataset collected from examinations at a Norwegian Hospital. Kvasir-Capsule consists of 117 videos which can be used to extract a total of 4,741,504 image frames. We have labelled and medically verified 47,238 frames with a bounding box around findings from 14 different classes. In addition to these labelled images, there are 4,694,266 unlabelled frames included in the dataset. The Kvasir-Capsule dataset can play a valuable role in developing better algorithms in order to reach true potential of VCE technology.

Gastrointestinal (GI) endoscopy has been an active field of research motivated by the large number of highly lethal GI cancers. Early GI cancer precursors are often missed during the endoscopic surveillance. The high missed rate of such abnormalities during endoscopy is thus a critical bottleneck. Lack of attentiveness due to tiring procedures, and requirement of training are few contributing factors. An automatic GI disease classification system can help reduce such risks by flagging suspicious frames and lesions. GI endoscopy consists of several multi-organ surveillance, therefore, there is need to develop methods that can generalize to various endoscopic findings. In this realm, we present a comprehensive analysis of the Medico GI challenges: Medical Multimedia Task at MediaEval 2017, Medico Multimedia Task at MediaEval 2018, and BioMedia ACM MM Grand Challenge 2019. These challenges are initiative to set-up a benchmark for different computer vision methods applied to the multi-class endoscopic images and promote to build new approaches that could reliably be used in clinics. We report the performance of 21 participating teams over a period of three consecutive years and provide a detailed analysis of the methods used by the participants, highlighting the challenges and shortcomings of the current approaches and dissect their credibility for the use in clinical settings. Our analysis revealed that the participants achieved an improvement on maximum Mathew correlation coefficient (MCC) from 82.68% in 2017 to 93.98% in 2018 and 95.20% in 2019 challenges, and a significant increase in computational speed over consecutive years.

Manual image annotation is a tedious and time-consuming task, while automated methods are error prone and limited in their results. Human computation, and especially games with a purpose, have shown potential to create high quality annotations by "hiding the complexity" of the actual annotation task and employing the "wisdom of the crowds". In this demo paper we present two games with a single purpose: finding regions in images that correspond to given terms. We discuss approach, implementation, and preliminary results of our work and give an outlook to immediate future work.

Having already discussed MatConvNet and Keras, let us continue with an open source framework for deep learning, which takes a new and interesting approach. TensorFlow.js is not only providing deep learning for JavaScript developers, but it's also making applications of deep learning available in the WebGL enabled web browsers, or more specifically, Chrome, Chromium-based browsers, Safari and Firefox. Recently node.js support has been added, so TensorFlow.js can be used to directly control TensorFlow without the browser. TensorFlow.js is easy to install. As soon as a browser is installed one is ready to go. Browser based, cross platform applications, e.g. running with Electron, can also make use of TensorFlow.js without an additional install. The performance, however, depends on the browser the client is running, and memory and GPU on the client device. More specifically, one cannot expect to analyze 4K videos on a mobile phone in real time. While it's easy to install, and it's easy to develop based on TensorFlow.js, there are drawbacks: (i) developers have less control over where the machine learning actually takes place (e.g. on CPU or GPU), that it is running in the same sandbox as all web pages in the browser do, and (ii) that in the current release it still has rough edges and is not considered stable enough to use in production.

Video games are often understood as engines of experience , and the interaction with the game lets players consume carefully constructed experiences. While it is generally agreed upon that a good experience makes a good game, methods for measuring or observing the impact of the gameplay on the players’ experience are still an open problem. In the 2021 Emotional Mario task, we ask researchers to investigate the gameplay of ten study participants on one of the most iconic classic video games: Super Mario Bros. We provide data to learn from, including heart rate, skin conductivity, videos of the players’ faces synchronized to the gameplay, the gameplay itself, and player demographics including their scores and times spent in the game. Participants of the task are asked to predict gameplay events based on the biometric and facial data of the players.

This paper describes the fourth edition of Lifelog challenges in ImageCLEF 2020. In this edition, the Lifelog challenges consist of two tasks which are Lifelog Moments Retrieval (LMRT) and Sport Performance Lifelog (SPLL). While the Lifelog Moments Retrieval challenge follows the same format of the previous edition, its data is a larger multimodal dataset based on the merger of three previous NTCIR Lifelog datasets, which contain approximately 191,439 images with corresponding visual concepts and other related metadata. The Sport Performance Lifelog, which is a brand new challenge, is composed of three subtasks that focus on predicting the expected performance of athletes who trained for a sport event. In summary, the ImageCLEF Lifelog 2020 receives 50 runs from six teams in total with competitive results.

Games are often defined as engines of experience, and they are heavily relying on emotions, they arouse in players. In this paper, we present a dataset called Toadstool as well as a reproducible methodology to extend on the dataset. The dataset consists of video, sensor, and demographic data collected from ten participants playing Super Mario Bros, an iconic and famous video game. The sensor data is collected through an Empatica E4 wristband, which provides high-quality measurements and is graded as a medical device. In addition to the dataset and the methodology for data collection, we present a set of baseline experiments which show that we can use video game frames together with the facial expressions to predict the blood volume pulse of the person playing Super Mario Bros. With the dataset and the collection methodology we aim to contribute to research on emotionally aware machine learning algorithms, focusing on reinforcement learning and multimodal data fusion. We believe that the presented dataset can be interesting for a manifold of researchers to explore exciting new interdisciplinary questions.

Writing source code for programs with lightweight text editors or fully featured integrated development environments is considered the main method of programming. Notebooks, however, are an extremely practical tool. In contrast to IDEs, projects are set up more easily and they allow for running programs in a read-eval-print loop (REPL) environment. The Jupyter Notebooks Quick Start Guide [1] describes notebook documents as "… both human-readable documents containing the analysis description and the results (figures, tables, etc..) as well as executable documents which can be run to perform data analysis." Basically, markdown text can be mixed with program source code in a sequence of sections, each dedicated to either programming or description and documentation. Source code sections can be executed and the output is appended to the section, even formatted in the form of graphs, diagrams, or tables. REPL and notebook based environments have proven to be useful in many scenarios including when exploring new libraries and frameworks, to prototype code, or as an educational tool to create interactive lecture material. A prominent example is Jupyter, which is the highly successful project behind Jupyter Notebooks and the recent JupyterLab, i.e. web-based systems to run and share notebooks that contain code, equations, data visualizations, and data exploration and narrative text.