CHUNG Wai-ching Bryan 鍾緯正 (HK|香港)

Movement in Space, part 2 (2018)

Computational installation with artificial neural networks generating a hologram of lines
Installation fixtures: Square table (1.5m X 1.1m height), 4 tablets on 4 tailormade plinths (1.1m), hologram projection, 3D animation
Digital equipment: Tablets – WIFI (27cm screen size); 1 computer with graphic card; 1 Wifi router; 4 micro-controllers (plug and unplug of wires); USB extension box with power
Floor space: 4 metre diameter
人工神經網絡繪製線條全息圖的運算裝置 (2018)
裝置:方桌(1.5米X 1.1米高),4個平板電腦 (放置於訂製底座) (1.1米),全息投影,3D動畫
數碼設備:平板電腦 –  無線網絡(27厘米屏幕尺寸); 1部電腦(配有顯卡); 無線網絡路由器; 4個微控制器(插拔電線); USB擴展盒 (配有電源)

03_test image_bryan_1.jpg


Movement in Space, part 2 (2018) is a re-interpretation of the former web-based project – Movement in Space, Experiments with Online Harmonograph. The work is an installation version with four users. It is an investigation of automatic drawing software, built from a number of harmonic motion equations. Each user can generate their drawing by using the graphical user interface from a tablet. The drawing will be a curve animated in a three-dimensional space. One user can also send the output from their drawing to another user, and in doing so, the drawings will grow more complex. The connections formed among the different users simulate the virtual connections from artificial neural networks. The Movement in Space project aims to research deep learning network applications in art and design, and to experiment with artificial neural network architecture in a creative art context.

Movement in Space, part 2 (2018) 是對藝術家以往基於網絡的作品-Movement in Space, Experiments with Online Harmonograph 的重新演繹。此作品為四個用家的裝置藝術版本,並且是對於一個由不同諧波運動方程組成的自動繪圖軟件的研究。 每個用家都可以使用平板電腦的繪圖用戶界面創作自己的繪圖-一個三維空間中的曲線動畫。 用家亦可以將他們的繪圖輸送給另一個用家,與此同時,圖像將隨之變得更加錯綜複雜。 不同用家之間所形成的連結將模擬人工神經網絡的虛擬連結。 Movement in Space 項目旨在研究深度學習網絡於藝術和設計上的應用,並在創意藝術的環境下對人工神經網絡架構作出實驗。


The project is an artistic investigation of an artificial neural network beyond its practical use in artificial intelligence (AI). An artificial neural network is a digital simulation of the connection network of human neurons, with applications mainly found in machine learning. With the popularity of the Google chess playing machine AlphaGo, the idea of AI has entered our everyday life conversation.

Traditional machine learning often requires large amounts of data for training. Through either supervised or unsupervised training, the machine, given new data, can perform its tasks of classifying and clustering the data or establishing relationships among the data. In the proposed project, the network will abandon its functional usage in machine learning and provide only visual responses as generative graphics. The project outcome is an online website where users can build a neural network graphically. Each neuron (node) in the network is a simple harmonograph capable of generating drawings with harmonic functions. By connecting the output from one node to the input of another, the data flow in the network can generate complex graphical forms.

The design of the network is a collaborative effort. Different users construct their own networks and connect to others. As a result, when online users are playing together at the same time, the activity becomes a live neural network with each user as an individual neuron. The drawings produced will be the emergent result of real time collaboration among the users. The feedback system is achieved by users observing their own drawings on screen; they can modify them as they go in real time, and thus also change the drawings of other users connected to them in a rippling manner.

The project provides an open-ended exploration of artificial neural network architecture. The drawings produced in the process may not be the major deliverables in the project. The co-drawing experience and the feelings of connectedness and mutually influencing each other are also the desirable outcomes of the project.

Movement in Space, part 2  has been rewritten from the original web version to a processing version. The animation is built with three parametric harmonic formulae. The outputs from one animation can be used as inputs for another formula, in order to simulate the artificial neural network.

From a game and play perspective, the Movement in Space series is an artistic investigation of automatic drawing machines, with reference to traditional drawing tools/toys such as the spirograph and harmonograph. The spirograph is a drawing toy for children to learn geometry through drawing; children with drawing pens follow gear movement to generate complex algorithmic drawings. The harmonograph was invented in 1844 by Professor Blackburn, originally for scientific use and later made into a drawing toy. Using a combination of pendulums, the machine is capable of drawing images similar to the wave form visualization generated from an oscilloscope.

In the project, the artist aims to simulate the mechanical movement of a pendulum in a harmonograph with the associated mathematical formulae from harmonic motion. In addition to the simulation, the project also explores the possibility of combining the loci of drawings from multiple users. In the end, it is no longer a drawing process, it can be a collaborative experience that encourages social interaction through drawing together. The drawing itself does deserve elaboration. A single user drawing may generate a drawing similar to what we expect from the wave form diagram from an oscilloscope. Multiple user drawings may incorporate complexity that is beyond anticipation. It will be a process of emergence.

Multiple users drawing may incorporate the complexity that is beyond anticipation. It will be a process of emergence.

Copy of test image_bryan_12.jpg

The basic formulae for each drawing are:

X = A x cos (B x t) + C x sin (D x t)

Y = E x sin (F x t) + G x cos (G x t)

where X and Y are the position of the drawing pen and A, B, C, D, E, F, G, H are the parameters that the user can choose to create their drawing. By routing the X and Y values to other users’ A to H, more complex forms can be generated.

Examples of drawings connecting more than one user
test image_bryan_203_test image_bryan_1Bryan_test image_bryan_3

The scenario does not stop here. When the software connects different users’ drawings, it can be a simulation of a neural network that behaves as human brain cells behave. Through the parametric control of the harmonic motion from other users’ drawings, it opens up possibilities to discuss feedback, memory, and cognition. The end result of the project will be an application or website that multiple users can connect together in real time to draw.

Implementation procedures of Movement in Space v. 2:

  1. Technical studies: to identify existing artificial neural network models and any corresponding software development kits, available in the open source software community. Since the majority of machine learning software libraries are only available in the Python programming environment, and which is not like the popular creative coding environment such as Processing or openFrameworks, the project will first research and determine the appropriate development and delivery platforms. (1 month)
  2. Interaction design: to design the front-end interface of the system. Ideally, it will be a web based platform where users can use ordinary web browsers to construct the network and produce the drawings. The project will also investigate the integration of various sensors in mobile device to enhance the drawing experience. (0.5 month)
  3. Application design: technical design of the web server component to implement the all the online communication and synchronization activities in the system. It involves the identification of reliable hosting service providers and the preparation of the infrastructure. (0.5 month)
  4. Software development: to implement both the front-end and back-end software. It also involves the extensive unit and integration test of system, across different front-end platforms. (6 months)
  5. Exhibition set-up: setting up and testing the site (0.5 month)

IT Elements 科技元素
Artificial neural network 人工神經網絡 / algorithmic animation 演算動畫 / microcontrollers 微控制器 /physical computing 物理計算


Dr. Bryan CHUNG is an interactive media artist and design consultant. His artworks have been exhibited at the World Wide Video Festival, Multimedia Art Asia Pacific, Stuttgart Film Winter Festival, Microwave International New Media Arts Festival and the China Media Art Festival. He was awarded the Hong Kong Arts Development Council’s Artist of the Year 2016 – Media Art, and the Grand Prize of the 19th Japan Media Arts Festival 2015, Art Division.

Chung once worked as a multimedia design consultant. In 2009, his consultation work, Coca-Cola Happy Whistling Machine won the Media Kam Fan Advertising Award. He has also provided consultancy to various industry leaders in Hong Kong and China for the former Shanghai Expo 2010. Chung studied computer science in HK, interactive multimedia in London, and software art in Melbourne. He also develops software libraries for the popular open source programming language Processing. He is the author of the books, Multimedia Programming with Pure Data (Packt Publishing, 2013) and Pro Processing for Images and Computer Vision with OpenCV (Apress Springer, 2017). Currently, he is Associate Professor in the Academy of Visual Arts, Hong Kong Baptist University, where he teaches subjects on interactive arts, computer graphics, and multimedia.

鍾緯正是互動媒體藝術家及設計顧問,作品曾在香港、北京、杭州、德國、荷蘭等地展出。其中《半百、半白》榮獲二零一五年日本媒體藝術祭,藝術組別大獎。二零零九年,他亦提供顧問及設計服務予《可口可樂、快樂工厰》項目,並奪得金帆獎媒體金獎。他在二零一零年上海世博期間為參展商提供互動設計顧問服務。鍾緯正在香港修讀電腦科學,其後在倫敦進修多媒體設計,並在澳洲墨爾本取得藝術博士學位。他亦為開源軟件 Processing 開發程式,及出版多媒體設計的教科書籍。目前他任教於香港浸會大學視覺藝術院,負責互動藝術、電腦圖像等科目。