Abstract

This paper presents the methodology employed for generating the video Anticipating 2020s which attempts to portray our collective imagination about how life in the 2020s would be from the perspective of decades-earlier science fiction films and books. The video was generated by rearranging the found footage of small excerpts from various science fiction movies set in the 2020s based on the output of a deep learning algorithm for object detection. Furthermore, Anticipating 2020s is subtitled with text generated by the deep learning model GPT-2 finetuned on the text of science fiction books set in the 2020s. The presented methodology explores the use of machine learning algorithms to create a database of found footage and the potential of trajectories through the database to create new narratives.

Abstract

Affective computing is an interdisciplinary field that studies computational methods that relate to or influence emotion. These methods have been applied to interactive media art-works, but they have been focused on affect detection rather than affect generation. For affect generation, computationally creative methods need to be explored that lately have been driven with the use of Generative Adversarial Networks (GANs), a deep learning method. The experiment presented in this paper, Forging Emotions, explores the use of visual emotion datasets and the working processes of GANs for visual affect generation, i.e., to generate images that can con-vey or trigger specified emotions. This experiment concludes that the methodology used so far by computer science re-searchers to build image datasets for describing high-level concepts such as emotions is insufficient and proposes utiliz-ing emotional networks of associations according to psychol-ogy research. Forging Emotions also concludes that to gener-ate affect visually, merely corresponding to basic psychology findings, e.g., bright or dark colors, does not seem adequate. Therefore, research efforts should be targeted towards under-standing the structure of trained GANs and compositional GANs in order to produce genuinely novel compositions that can convey or trigger emotions through the subject matter of generated images.

Abstract

One of the contributing factors to the continuing debate among art historians over the use of computational methods in art history research is that they do not consider the core of today's art history research questions. The lack of close collaboration between the two involved research communities makes the definition of contemporary art-historical methods as well-defined computer vision problems extremely difficult. For that purpose, it is devised as a methodology to study articles in art history journals from a computer science perspective. The objective is to identify which image features art historians utilise within their research and describe them in immediate and meaningful terms to the computer vision research community. Finally, some paradigms that could serve as a new starting point for exploring how computer vision applications for art history can address the core of today's art history research are given.

Abstract

Breaking the Silence uses machine learning methods to understand what are the contemporary taboo topics. Initially, news articles that contained the word taboo were mined from various online sources. The mined articles were originally posted from January 2019 to September 2020. These articles include overt references to contemporary taboo topics in today’s press as well as the report of transgressions. Natural Language Processing (NLP) methods are then used to identify the main topic of each mined article and thus group them into sets according to similarity and relevance. Next, techniques for information extraction are used to determine the most meaningful elements of the articles’ text like entities, abstract concepts, etc. Finally, new documents are generated with the utilization of automatic text summarization techniques. These generated documents summarize the main ideas and opinions for each identified taboo topic. The images shown in all articles were generated with Runaway Generative Engine by randomly selecting sentences from each article. Breaking the Silence aims to highlight and bring to focus what topics are currently discussed as taboos. The generated summarizations underline the different topics discussed at different times and additionally the arguments for the creation, maintenance, or destruction of a taboo. Visit the Breaking the Silence website.

Abstract

The Invisible Structures of the Artworld draws inspiration from the contemporary artworld processes through which artists and artworks are appreciated and gain recognition as well as the fact that the artworld structure at any given time is not entirely known. Another point from which this work draws inspiration is the approaches taken for social network analysis, the process of investigating social structures through the use of networks. The Invisible Structures of the Artworld is a data-driven work. It collects real-time social media data to identify artworld actors and finally visualize an artworld network whose structure becomes visible. Thus, it visualizes the global transformations of the contemporary artworld as well as the role and impact of artworld actors involved in the processes of the production and mediation of art.

The job description of an engineer, of any specialty - civil, electrical, mechanical, chemical or computer engineer, is common: to solve problems, face challenges and problems of the real world with creative and innovative solutions aiming to make the world work better and contribute to progress. In the context of this paper, we will consider the forming of a curriculum for a 21st century School of Fine Arts to be a “problem”. The logic and process of problem-solving will be followed in the same manner as it would be followed in a School of Engineering. This approach will allow us to identify and study the different perceptions and practices in art and engineering education. As we would do for solving any other problem, we will first examine the situation from which we start, the input, and the situation in which we want to conclude, the output. In our case, we will initially examine the reasons for which someone enters a school, their expectations from it and eventually the knowledge, skills, and abilities that they are expected to have as graduates of this school to accomplish a successful career path. The study of the defined problem of forming a curriculum will be tackled with a systems design approach that allows us to move from an initial to a final desired state. This will lead us to identify the key differences and/or similarities in the two kinds of Schools that are examined. More importantly, it will allow us to explore the different ways in which a system like that can be designed, while, at the same time, it will be juxtaposed with the existing curricula and teaching methods in Schools of Fine Arts.

Abstract

Time series prediction involves the determination of an appropriate model, which can encapsulate the dynamics of the system, described by the sample data. Previous work has demonstrated the potential of neural networks in predicting the behaviour of complex, non-linear systems. In particular, the class of polynomial neural networks has been shown to possess universal approximation properties, while ensuring robustness to noise and missing data, good generalisation and rapid learning. In this work, a polynomial neural network is proposed, whose structure and weight values are determined with the use of evolutionary computing. The resulting networks allow an insight into the relationships underlying the input data, hence allowing a qualitative analysis of the models' performance. The approach is tested on a variety of non-linear time series data.

Abstract

This paper proposes a novel hierarchical representation of POMDPs that for the first time is amenable to real-time solution. It will be referred to in this paper as the Robot Navigation - Hierarchical POMDP (RN-HPOMDP). The RN-HPOMDP is utilized as a unified framework for autonomous robot navigation in dynamic environments. As such, it is used for localization, planning and local obstacle avoidance. Hence, the RN-HPOMDP decides at each time step the actions the robot should execute, without the intervention of any other external module. Our approach employs state space and action space hierarchy, and can effectively model large environments at a fine resolution. Finally, the notion of the reference POMDP, that holds all the information regarding motion and sensor uncertainty is introduced, which makes our hierarchical structure memory efficient and enables fast learning. The RN-HPOMDP has been tested extensively in a real-world environment.

Abstract

This paper introduces a methodology for avoiding obstacles by controlling the robot's velocity. Contemporary approaches to obstacle avoidance usually dictate a detour from the originally planned trajectory to its goal position. In our previous work, we presented a method for predicting the motion of obstacles, and how to make use of this prediction when planning the robot trajectory to its goal position. This is extended in the current paper by also using this prediction to decide if the robot should change its speed to avoid an obstacle more effectively. The robot can choose to move at three different speeds: slow, normal and fast. A hierarchical partially observable Markov decision process (POMDP) controls the robot movement. The POMDP formulation is not altered to accommodate for the three different speeds, to avoid the increase of the size of the state space. Instead, a modified solution of POMDPs is used.

Abstract

This paper considers the problem of a robot navigating in a crowded or congested environment. A robot operating in such an environment can get easily blocked by moving humans and other objects. To deal with this problem it is proposed to attempt to predict the motion trajectory of humans and obstacles. Two kinds of prediction are considered: short-term and long-term. The short-term prediction refers to the one-step ahead prediction and the long-term to the prediction of the final destination point of the obstacle's movement. The robot movement is controlled by a partially observable Markov decision process (POMDP). POMDPs are utilized because of their ability to model information about the robot's location and sensory information in a probabilistic manner. The solution of a POMDP is computationally expensive and thus a hierarchical representation of POMDPs is used.