SIGGRAPH Asia 2019: Deep Dreaming

• Hector Rodriguez

This talk explores some of the possible applications of unsupervised machine learning methods in found footage cinema, a tradition of experimental art that re-edits excerpts from existing films. This artistic practice sometimes aims to reconfigure our experience of the moving image heritage. In this context, machine learning algorithms has the potential to capture aspects of the cinematic experience for which we lack critical concepts, and which are for this reason difficult to describe. One important example concerns cinematic motion. Established critical discourse often speaks of motion in film by reference to the movement of objects or the camera. Film scholars might describe a scene by noting, for instance, that a person is walking fast or that the camera is tilting upwards. What is missing in this kind of description is the visual texture of cinematic movement. The two-channel algorithmic installation Errant: The Kinetic Propensity of Images applies matrix factorization techniques to the analysis of optical flow in cinema, focusing on the work of Chinese director King Hu. This method produces a visual dictionary of basic motion patterns the represent what could be described as the “kinetic overtones” of image sequences. The results are then visualized using streaklines, a technique from fluid dynamics. This presentation will discuss the motivation and methodology used in the production of this work, in relation to other work by the speaker. Implications for cinema theory will also be briefly discussed.

• Santiago Echeverry

Anathema is the psychological study of Doug’s demons, nightmares and fantasies in a dream-like autoscopy that displays the strengths and vulnerabilities of a self-altered and self-made man in his 70’s. The piece was created with the Kinect volumetric sensor, with music by Mexican cyber-punk musician Cesar Cardenas aka Zoonosis.

• Jon McCormack

With the recent renewed interest in art generated by Artificial Intelligence (AI), it is timely to re-explore the body of knowledge and critique around art made by algorithms. Since computers were first adopted as art machines a number of enduring criticisms have reoccured over the decades, often with different names, but ultimately similar conceptual foundations. Essentially they relate to issues of authorship (who is the author when an “intelligent” machine is involved in the art making process?), autonomy (how much of the decision making and creative judgment is absolved to the machine?), authenticity (can creative acts or outputs made by machines ever be authentic?) and intention (is it right to think of machines as artists?). As AI technologies are increasingly fetishised by technologists and artists, a renewed debate around these criticisms has reemerged. In this talk I want to specifically address the issue of algorithmic genericism: how can algorithmic art practices escape the spectre of being generic to the algorithm itself? How can a practice be informed to recognise the issues of authorship, autonomy, authenticity and intention and move beyond algorithmic genericism?

• Kit Devine

Heritage visualisations are works of the cultural imaginary and this paper examines the artwork Artistic License: VR Sydney Cove ca. 1800 which foregrounds the interpretive nature of heritage visualisation. It is a re-imagining in virtual reality of A View of Sydney Cove, New South Wales, 1804, a contemporaneous print of Sydney Cove. Existing in the liminal space between accuracy and authenticity it is both art object and heritage visualisation. The dual nature of this work supports engagement with wider audiences, fostering and broadening debate at individual, institutional, academic and societal levels about the nature and role of heritage.

• Peng Tan
• Yi Ji
• Damian Hills
• Tieming Fu

The purpose of this paper is to explore new perspectives to learning Intangible Cultural Heritage (ICH) through embodied interaction with focus on learning and experience with traditional Cantonese Porcelain crafting. This research developed a WebAR application where various processes are presented through the tangible interaction of virtual porcelain represented by physical objects. The learner is able to directly interact with the plate that bridges the tangible materials and making processes of ICH utilizing WebAR. Empirical studies found that the WebAR and embodied interaction can enhance student’s tangible learning experience to transfer knowledge between craftsman and student.

• Sophia Brueckner

Captured by an Algorithm is a commemorative plate series that looks at romance novels through the lens of the Amazon Kindle Popular Highlight algorithm. Each plate features one highlight and an algorithmically generated landscape. The highlights tell a story of the loneliness, grief, vulnerability, and discontent felt by the readers.

• Masaru Mizuochi

The personality of living things is beautiful. Is it possible to have a personality in the digital world? CharActor is a video work that produces animation by recognizing shader programs as genes and by changing the mathematical expression itself using evolutionary computation.

• Rewa Wright
• Simon Howden

When we feel and sense through machines, are we still ourselves? In a mixed reality where embodied actions and blinding visions are part woman/part machine, the tactile surface of plants is a portal that conjures augmented materialities into existence.

• Baden Pailthorpe
• Charles Gretton
• Rhys Healy

Dream Clanger is a hybrid art/computer science project that re-imagines AFL Player GPS data and match video. Building on Baden Pailthorpe’s 2017 major exhibition ‘Clanger’, this work pushes the envelope further by integrating machine learning.

• Tomás Laurenzo
• Alejandro Rodríguez
• Tatjana Kudinova

This paper presents Extraordinary Accident, an immersive experience exploring how different levels of abstraction can coexist and collaborate in a representation and recreation of urban space. Using Hong Kong as both inspiration and data source, the work attempts to liberate virtual reality compositions from their metaphorical ballast –that is, their recreational onus– and instead, with a temporal amalgamation of poetic representation at different scales, contribute to an alternative, potentially more intimate, understanding of the urban experience.

• Suguru Goto

This paper reports the exposition of an artistic installation, gravityZERO, and its ongoing technical development. It consists of virtual sound, VR and robotic technologies in order to simulate the state of zero gravity. Audience members can experience a floating sensation within this virtual environment.

• Simon Rippingale
• Andrew Johnston
• Andrew Bluff

Through a detailed account of a recent practice-based research project – a short animation project called Jasper, this paper explores how a hybrid analogue/digital production approach can generate a unique and engaging visual style – one that sits between the tangible, handcrafted feel of miniatures and the cleanness, fluidity and flexibility of computer-generated animation. The author examines the new creative possibilities and challenges that a hybrid animation production approach presents and also outlines various technical platforms encountered during the production of Jasper, including motion-controlled camera systems, 3D printing, game engines, point cloud scans and augmented reality.

• Haein Kang

ILLUSION explores the relationship between the body, mind, and machine by taking advantage of the brain-computer interface. It detects whether your consciousness has visual stimuli to produces cadenced sound, an exterior manifestation of the performer’s internal state. When you close your eyes, you can see the world imaged by sounds.

• Emi Kusano
• Junichi Yamaoka

This artwork is an installation that expresses the future in which users can manufacture designer’s babies themselves. You can design, customize and manufacture your baby with your favorite gene on your laptop. A 3D printed child appears from the display, and the child’s face created based on the visitor’s face.

• Uwe Rieger
• Yinan Liu

LightWing II creates a mysterious sensation of tactile data. In this interactive installation, a kinetic construction is augmented with stereoscopic 3D projections and spatial sound. A light touch sets the delicate wing-like structure into a rotational oscillation and enables the visitor to navigate through holographic spaces and responsive narratives.

• Refik Anadol

Machine Hallucinations – Latent Study II is part of an ongoing synthetic reality collection that explores the relationship between memory and dreams, recognition and perception.

• Andrew Yip

Metascape: Villers Bretonneux is an immersive, interactive memoryscape experienced in first-person perspective, that simulates 72 hours in real time of the 1918 First World War Second Battle of Villers Bretonneux. The work relies on multiple forms of spatial and memory reconstruction, both driven by algorithmic processes.

• Øyvind Sørdal Klungre
• Asim Hameed
• Andrew Perkis

Our ways of telling stories have evolved along with advances in technology. This has led to the emergence of digital storytelling, emphasizing multimodality and interactivity. This project explores narrative influences on Quality of Experience of users in digital stories. This is done by creating and implementing a location driven digital story presented to the user by an augmented reality application made in Unity on a mobile device. This narrative system has then been evaluated by 30 people who have participated in a subjective evaluation. The results show that the narrative setup results in a richer, livelier and more engaging experience.

• Taeil Lee

Numb, shaped as an exaggerated eyeball, follows you and reflects the blinks of yours. It makes you become aware of your own blinking and sensitive to your own sensation. Numb illustrates how we build relationship with technology through senses, and how we become sensitive to ourselves by and with technology.

• Clarissa Ribeiro
• Herbert Rocha

Combining algorithmic modeling strategies for data visualization with digital fabrication, the work includes the generative design of a series of geometrically intricate crystals-like 3D models using as raw data Twitter APIs having as the search phrase hashtags related to Brazilian 2018 presidential elections twitted from defined geolocations.

• Tomás Laurenzo

Smile is a mixed-media installation consisting of a screen in a black box, mounted on the wall. When an interactor smiles, drone-footage of the ruins of Gaza fades in. If the interactor stops smiling, the video stops. It only plays when the interactor widely smiles at it.

• Dennis Del Favero
• Tomasz Bednarz

SoS recreates the experiences of two Syrian asylum seekers as they lose sight of each other during a treacherous ocean voyage from Indonesia to Northern Australia.

• Toshikazu Ohshima

Tactile Microcosm of ALife offers interaction with artificial organisms, whereby the user can enjoy playing with fish-like organisms through aerial imaging and haptic feedback. The holographic organisms float in water in a petri dish, and the user can feel a forcefield of the vital of the organisms via force feedback.

• Anju Niwata
• Hidenori Watanave

We explain the creation of “flow” in social media and real spaces using AI technology to colorize black-and-white stock photographs from digital archives and other locations. When visualizing the colors that the photographs should have had, the impressions of “freezing” in black-and-white photographs are “rebooted,” and viewers can more easily imagine the events depicted. This bridges the psychological gap between past events and modern daily life, sparking conversations. The “flow” generated here causes the emergence of lively communication and increases the value of information. This method can help to pass precious materials and memories of past events into the future.