Conference PaperPDF Available

Generative Design of Urban Fabrics Using Deep Learning


Abstract and Figures

This paper describes the Urban Structure Synthesizer (USS), a research prototype based on deep learning that generates diagrams of morphologically consistent urban fabrics from context-rich urban datasets. This work is part of a larger research on computational analysis of the relationship between urban context and morphology. USS relies on a data collection method that extracts GIS data and converts it to diagrams with context information (Rhee et al., 2019). The resulting dataset with context-rich diagrams is used to train a Wasserstein GAN (WGAN) model, which learns how to synthesize novel urban fabric diagrams with the morphological and contextual qualities present in the dataset. The model is also trained with a random vector in the input, which is later used to enable parametric control and variation for the urban fabric diagram. Finally, the resulting diagrams are translated to 3D geometric entities using computer vision techniques and geometric modeling. The diagrams generated by USS suggest that a learning-based method can be an alternative to methods that rely on experts to build rule sets or parametric models to grasp the morphological qualities of the urban fabric.
Content may be subject to copyright.
1,2Computational Design Laboratory, CRAIDL, Carnegie Mellon
Abstract. This paper describes the Urban Structure Synthesizer (USS),
a research prototype based on deep learning that generates diagrams
of morphologically consistent urban fabrics from context-rich urban
datasets. This work is part of a larger research on computational analysis
of the relationship between urban context and morphology. USS relies
on a data collection method that extracts GIS data and converts it to
diagrams with context information (Rhee et al., 2019). The resulting
dataset with context-rich diagrams is used to train a Wasserstein GAN
(WGAN) model, which learns how to synthesize novel urban fabric
diagrams with the morphological and contextual qualities present in the
dataset. The model is also trained with a random vector in the input,
which is later used to enable parametric control and variation for the
urban fabric diagram. Finally, the resulting diagrams are translated to
3D geometric entities using computer vision techniques and geometric
modeling. The diagrams generated by USS suggest that a learning-based
method can be an alternative to methods that rely on experts to build
rule sets or parametric models to grasp the morphological qualities of
the urban fabric.
Keywords. Deep Learning; Urban Fabric; Generative Design;
Artificial Intelligence; Urban Morphology.
1. Introduction
Urban fabric is a key concept in urban design that consists of the configuration
of streets, parcels, lots, and buildings (Oliveira, 2016, p. 8). It operates
in the intersection of urban and architectural scale and can reflect important
aspects of social phenomena and cultural practices. Given the importance of
the configuration of the urban fabric in its inhabitants’ lives, understanding,
categorizing, and designing them has been a long-lasting challenge for design
In the past decades, there have been several studies on the computational
synthesis of urban fabrics using different computational methods, such as
multi-agent system (Biao et al., 2008), diffusion-limited aggregation (Koenig,
2011), rule-based system (Pellitteri et al., 2010), L-systems (Parish & Müller,
n.d.), etc. While these computational methods can generate good results, they
PROJECTIONS, Proceedings of the 26th International Conference of the Association for Computer-Aided
Architectural Design Research in Asia (CAADRIA) 2021, Volume 1, 31-40. © 2021 and published by the
Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Hong Kong.
are limited to (1) modes of operation of existing generative models and (2) the
designers’ capacity to understand and encode the desired morphological qualities
and contextual adaptations in principles, rules, and parameter calibration of the
In this research, we present Urban Structure Synthesizer (USS), a
learning-based model to synthesize urban fabric diagrams that alleviate the
reliance on knowledge explicitly embedded in generative models. Unlike
rule-based or agent-based system, which depends on experts to create rules for
generation, USS learns the morphological features of urban fabrics from data.
Firstly, we use a method to convert site data from GIS into diagrammatic
images with contextual information introduced in the previous research (Rhee et
al., 2019). After curating the site data with the desired patterns of urban forms and
features, we use the resulting context-rich dataset to:
train an analytical model that organizes the data into clusters with similar
morphological types (Rhee et al., 2019).
train a WGAN model to synthesize site-responsive diagrams of urban fabric
based on the morphological qualities of the sites in the dataset (Figure 1).
Figure 1. Two different processes: analysis (top) and generation (bottom) of urban fabrics
using different machine learning models. USS interface connects the analytical and generative
2. Learning
2.1. DATA
USS employs a diagrammatic image dataset (DID), a data synthesis technique
for raster images with a diagrammatic representation of two-dimensional building
form and its neighboring urban contexts. Diagrammatic image representation in
the dataset “offers two advantages in machine learning: low level of image noise
and support for custom selection of morphological information”.(Rhee et al., 2019,
p. 345).
Currently, the most popular way to create images of urban information is by
using satellite and map images. Both methods can add unnecessary information or
noise to the images. Therefore, a method such as image segmentation is typically
employed to remove the noise of the representation for datasets based on satellite
images. In contrast, diagrammatic image representation is synthesized from GIS
information selected by the user, which preemptively removes or even excludes
image noise from the dataset.
Figure 2. Composition of DID-PGH dataset in three different level.
Taking advantage of these qualities, we use DID to create a dataset of
Pittsburgh, USA, which we will referreferred to as DID-PGH. DID-PGH
comprises images with a target building placed on the center, and their neighboring
building footprints, street network, and the parcel shapes around. The height of
center building is represented through the color of the solid inside the polygon of
the footprint. The more reddish color indicates that the building is lower, while
the more yellowish color indicates that the building is higher. Each image is 512
x 512 pixels, and the dataset has a total of 45,852 images (see previous research
for more information, Rhee et al., 2019).
The selection of the generator for USS was based on an initial test of three
different generative deep learning models for urban fabric synthesis: Variation
Autoencoder (VAE), Generative Adversarial Networks (GAN), and Wasserstein
Generative Adversarial Networks (WGAN). We compared the resulted images
from each model after training all models with the context-rich dataset in the same
training epochs. VAE resulted in the blurriest images. The images from GAN was
sharper than VAE’s one, but it hardly captures the morphological features form
the context-rich dataset. With more stable training, the images created by WGAN
tend to be sharper than the images generated by the simple GAN.
USS specifically uses Wasserstein Generative Adversarial Networks - Gradient
Penalty (WGAN-GP) (Gulrajani et al., 2017) as the model to generate new urban
fabric diagrams based on the captured patterns of contextual and morphological
features of the dataset. The overall structure of this model is similar to other
Generative Adversarial Networks (GAN) model, except for some changes in the
components of the optimization, such as the error function and the gradient penalty
term. It has two networks: a generator and a critic. During training, the generator
creates images conditioned on random input vectors, and the critic provides a value
signal indicating how real each input image is. In the initial step of training, the
generator synthesizes random images, and the critic provides random signals. The
key to GAN lies in how we alternate the training of the two networks so that the
generator becomes more adept at fooling the critic and the critic at identifying
which observations are fake (Foster & Safari, 2019, p. 100).
Figure 3. The structure of WGAN-GP model for training DID-PGH dataset.
In this research, the model generates a new urban fabric by learning patterns
of contextual features of a given building. The generator of WGAN-GP creates an
urban fabric image using a noise vector of size 100 as the input parameters.
Figure 4. Comparison of images generated by different deep learning models.
We trained this model for about 10000 batches (15 epochs) with 64 for batch
size. The model’s optimizer is ‘Adam’ (Kingma & Ba, 2017), and the learning rate
is 2.0E-4. The model has no dropout layers, and both discriminator and generator
use Leaky Rectified Linear Unit (Leaky ReLU) as the activation function. This
model was trained on a computer with the following specifications: ‘Intel(R) Core
(TM) i7-8700k @ 3.70GHz’, 64GB memory, and two GTX-1080ti graphic cards.
It took almost 37 hours to train the model. The losses reduced significantly until
3000 batches, , and after that, the changes were subtle.
Figure 5. Changes of loss and image quality during the training process.
3. Design Implementation
To generate urban fabric configurations, USS has an interface that relies on
three steps: loading the trained model in the back-end, communication between
front-end and back-end, and object tracking system.
When the USS is launched, the trained model is loaded in the back-end of
the interface. Users can see an image of the urban fabric, sampled sliders, and
buttons in the front-end fo the interface. The sliders represent the values of the
noise Z vector for image synthesis. After training, each value of the vector can
capture a design variation consistent with the design space of the original dataset.
From the users’ perspective, these sliders can be used to control morphological
variations of the synthetic urban fabric in real-time. Although the original size
of the vector is 100, we sampled 31 values that are represented as sliders in the
interface. After manually changing the values of the sliders, the users can click on
the ’GENERATE’ button to pass the updated noise vector to the generator, which
returns an updated synthetic image.
Figure 6. Interface of USS.
Figure 7. USS user interface system configuration.
To store the results, USS extracts the polylines of buildings and using OpenCV
- a library for real-time computer vision. Each detected and tracked 2D information
is reconstructed into 3D information using Rhinoceros®, the popular modeling
tool in architectural design, and Grasshopper, a graphical algorithm editor tightly
integrated with Rhinoceros.
Through custom components in Grasshopper, USS can communicate with
Grasshopper in real-time and users can import the tracked object’s outlines and
its bounding boxes. After importing, users can see automatically reconstructed a
3D urban fabric based on the synthesized image from USS. USS tracks the color
of the center building,converts the color value into height value according to the
preset of color range, and saves this value so that users can import this height
information in Grasshopper to build a 3D model.
In order to test USS, we propose a speculative design exercise in Pittsburgh. The
design uses USS to insert urban elements based on the morphological qualities of
downtown Pittsburgh into the typical low-rise residential area of Shadyside. This
exercise delves into the potential of how USS can be used for speculating about
conflicts of scale and morphology.
Figure 8. The changes in urban fabric images according to the 25th and 11th feature slider
Firstly, a urban fabric grid is applied to the site. The tiles of the grid have
the same range and size as the image to be synthesized by USS. Then, part of the
existing urban fabric is removed from several grid tiles to secure the place for new
Figure 9. Urban fabric design process using USS.
Secondly, the downtown fabric tiles are customized, produced, and displayed
in real-time based on slider values of the interface. Each slider captures different
functional features. For example, the 11th slider influences the orthogonal
arrangement of buildings and streets. The 25th slider controls the merging of
buildings in the bottom center of the tile.
Designing an urban tile with sliders in USS consists of two main steps: global
and local setting. The global setting loads a synthesized image of the desired urban
fabric through presets. The slider values will be changed according to the presets.
The local setting contains adjusting the preset sliders for fine-tuning the global
While the user changes sliders, USS uses the analytical model to show the
types of urban fabrics that the currently generated image is most similar to. The
analytical model becomes a visual guide that indicates how users can modify
the global and local settings to design the intended fabric. When users finished
modifying the sliders and press to confirm button, USS places the reconstructed 3D
model of the fabric on the site. Repeating these adjusting and confirming process
eight times, we have generated eight downtown-like urban fabric tiles.
USS is used to transplant the synthesized downtown urban fabric to the
residential area of Shadyside. The downtown tiles are placed without rigid
borders between the existing and newly proposed fabrics. Therefore, the existing
residential grid was replaced by the downtown business grid, which meant that
the proposed fabrics had to be re-aligned with the residential fabrics. Through this
design process, a downtown-like organization was created in the given area.
The speculative design presented in the previous section resulted in a
heterogeneous fabric at the center of the urban residential site in terms of
building sizes, shapes, and the axis of the grid with the gradual changes at the
periphery. The buildings in the proposed urban fabric are larger and higher than
the buildings in the previous fabric. The grid axes of the previous fabric are almost
orthogonal and form rectangular grid patterns. On the other hand, the proposed
fabric has diagonal axes that form triangular grid patterns. This resulting urban
collage exposes the contrast between the different morphologies with an intricate
geometric model generated in real-time.
Figure 10. Comparison of the previous and proposed urban fabrics.
Figure 11. Different form of urban space in the previous and proposed urban fabrics.
Figure 11 shows the difference between the previous residential urban fabrics
and the proposed business urban fabrics in detail. In cases A and B, the proposed
fabric designed through USS has different size of buildings, clusters, and blocks,
but keeps the existing street system and its axes. Cases C and D show that the
streets in a new direction were created by utilizing characteristics of the existing
fabrics. In the areas where different fabrics conflict, the periphery of the fabrics
became indistinct by gradually changing the size of the buildings and directions
of streets.
Our experiment shows the potential of USS to support a new urban
design approach based on the investigation of urban fabric designs that are
morphologically consistent. For example, like the experiment above, designers
could create another downtown variation with multi-core structure (instead of
uni-core) by preserving a the global downtown setting and exploring local
solutions with the slides. Instead of explicitly modeling streets, building blocks,
urban amenities, and public space, urban designers or planners can interact with
real-time variations of fabric morphology on the site for design ideation.
4. Conclusion
USS illustrates the potential of using deep learning model in the generation of
urban fabric, which enables designers not only to get access to the statistical
features of datasets but also to produce novel designs from these features.
Through deep learning, it was possible to design a new urban fabric by
extracting complex patterns from the existing fabrics and synthesizing the patterns
according to a dataset curated by the designer. The deep learning-based design
system is an alternative to generative systems that require human expertise to
access and set generative rules and parameters. Learning-based systems can
handle larger amounts of and more complex features than hand-craft rule-based
and parametric systems. For example, in the design example of this study, the
user did not define the rules of how to generate and position the new urban fabric.
The deep learning model synthesized a new fabric by learning the pattern of the
numerous existing fabrics. The learned function contains transformations that are
potentially more complex and adaptive than rules explicitly defined by experts.
Models based on more intricate rules can reveal design patterns and knowledge
embedded in the built environment and, therefore, can support designers to devise
new design realms.
One of the main limitations of USS is the lack of control over the results.
By relying on a random vector, the control of the variations of the result
is not easy to grasp by the users. In order to extend USS to real-design
applications, it is necessary to explore useful and reliable human-computer
interactions and control mechanisms. There are some interesting alternatives from
the generative modeling perspective, such as using paired datasets for a conditional
generation. Considering that the dataset used in USS is already pre-processed,
it is straightforward to derive custom diagrammatic controls such as the axes,
masses (Zheng, 2018), connections in the border, etc. From the human-computer
interaction, it is necessary to establish systematic studies with the users to identify
adequate interaction modes with design variations.
Biao, L., Rong, L., Kai, X., Chang, L. and Qin, G.: 2008, A GENERATIVE TOOL BASE ON
MULTI-AGENT SYSTEM, Proceedings of the CAADRIA 2008.
Foster, D.: 2019, Generative Deep Learning : Teaching Machines to Paint, Write, Compose,
and Play, O’Reilly Media.
Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V. and Courville, A.C.: 2017, Improved
Training of Wasserstein GANs, Advances in Neural Information Processing Systems 30,
Kingma, D.P. and Ba, J.: 2017, Adam: A Method for Stochastic Optimization,
Koenig, R.: 2011, Generating urban structures: A method for urban planning supported by
multi-agent systems and cellular automata, Przestrzeń i Forma (space & FORM),16,
Oliveira, V.: 2016, Urban Morphology: An Introduction to the Study of the Physical Form of
Cities, Springer International Publishing.
Parish, Y.I.H. and Müller, P.: 2001, Procedural Modeling of Cities, Proceedings of the 28th
Annual Conference on Computer Graphics and Interactive Techniques, New York, 301–308.
Pellitteri, G., Lattuca, R. and Conti, G.: 2010, A Generative Design System to Interactively
Explore Different Urban Scenarios, Proceedings of the eCAADe2010.
Rhee, J., Cardoso Llach, D. and Krishnamurti, R.: 2019, Context-rich Urban Analysis Using
Machine Learning: A case study in Pittsburgh, PA, Proceedings of the 37th eCAADe and
23rd SIGraDi.
Zheng, H.: 2018, Drawing with Bots: Human-computer Collaborative Drawing Experiments,
Proceedings of CAADRIA 2018.
... For example, Wu et al. [27] conducted a comprehensive review of how GANs (generative adversarial networks) are applied to tackle challenging tasks in the built environment and identified 26 unique application domains enabled by GAN including floorplan generation, street-view generation, satellite image enhancement, etc. In terms of urban planning and building design, a few pioneering studies have started to apply deep generative methods including GAN and VAE (variational autoencoder) for automated building layout generation since 2020 [3,[27][28][29][30][31][32][33]. In contrast to rule-based methods and evolutionary algorithms, deep generative methods are developed based on deep neural networks to learn hidden relationships and urban knowledge from real-world design cases and synthesize visually realistic and semantically reasonable building solutions. ...
Generative adversarial network (GAN) Design scenario 3D visualization A B S T R A C T Building layout generation has entered a new era in recent years, leveraging state-of-the-art deep generative methods to learn morphological properties of exiting urban structures and synthesize building alternatives responsive to local context. However, most existing research generally follows an image-to-image translation idea, while overlooking the impact of site/design attributes on building configuration, making their results less performative. Besides, most synthesized layouts are commonly displayed in 2D pixelized images, limiting further performance evaluation and informed decision-making. This study, therefore, proposes a novel GAN-based model, namely site-embedded generative adversarial networks (ESGAN) for automated building layout generation. Both qualitative and quantitative results in New York City indicate ESGAN is capable of synthesizing visually realistic and semantically reasonable layouts. This end-to-end generative system can not only encode a conditional vector to improve performance in different design scenarios but also display synthesized layouts at different levels of detail for human-system interaction.
... 29 Not least, pattern recognition with deep understanding enables us to investigate inductive morphological gradients of urban and architectural forms. 30 These are just a few examples of how artificial intelligence can advance computational design as a discipline. This list is by no means complete. ...
Conference Paper
Full-text available
This paper explores the optimization of apartment building layouts using five different meta-heuristic optimization algorithms: Simple Genetic Algorithm (SGA), Simple Evolutionary Algorithm (SEA), Self-Adaptive Differential Evolution (SADE), Particle Swarm Optimization (PSO), and Artificial Bee Colony (ABC). The study aims to compare and analyze the performance of each algorithm in solving the apartment building layout problem. The optimization process involves placing apartment buildings on a given site while considering design constraints such as floor area ratio, building separation distance, and building height. The objective function is formulated to minimize the number of buildings violating these constraints. The results show that SGA outperforms the other algorithms, achieving the lowest error rate and producing better layouts. SADE follows with a relatively slower but effective optimization performance. SEA and ABC exhibit similar patterns, while PSO demonstrates the least satisfactory results, indicating the need for further optimization. The study provides valuable insights into the characteristics of each algorithm in apartment building layout optimization and can serve as a basis for future research in this field.
Urban design is the process of designing and shaping the physical forms of cities, towns, and suburbs. It involves the arrangement and design of street systems, groups of buildings, public spaces, and landscapes, to make the urban environment performative and sustainable. The typical design process, reliant on manual work and expert experience has unavoidable low efficiency in generating high-performing design solutions due to the involvement of complex social, institutional, and economic contexts and the trade-off between conflicting preferences of different stakeholder groups. Taking advantage of artificial intelligence (AI) and computational capacity, generative urban design (GUD) has been developed as a trending technical direction to narrow the gaps and produce design solutions with high efficiency at early design stages. It uses computer-aided generative methods, such as evolutionary optimization and deep generative models, to efficiently explore complex solution spaces and automatically generate design options that satisfy conflicting objectives and various constraints. GUD experiments have attracted much attention from academia, practitioners, and public authorities in recent years. However, a systematic review of the current stage of GUD research is lacking. This study, therefore, reports on a systematic investigation of the existing literature according to the three key stages in the GUD process: (1) design problem formulation, (2) design option generation, and (3) decision-making. For each stage, current trends, findings, and limitations from GUD studies are examined. Future directions and potential challenges are discussed and presented. The review is highly interdisciplinary and involves articles from urban study, computer science, social science, management, and other fields. It reports what scholars have found in GUD experiments and organizes a diverse and complicated technical agenda into something accessible to all stakeholders. The results and discoveries will serve as a holistic reference for GUD developers and users in both academia and industry and form a baseline for the field of GUD development in the coming years.
Full-text available
Nolli map is the earliest diagram tool to simplify and quantify urban form, which most intuitively reflects the spatial layout of tangible elements in the city. The urban morphology contains its inherent evolutionary laws. Exploring the inner rules of cities is helpful for people to conduct urban research and design. Unlike the traditional research methods of urban morphology, the neural network algorithm provides us with new ideas for understanding urban morphology. In this experiment, we label 136 European cities samples in the rules of Nolli map as a training set for machine learning. We use Generative Adversarial Networks (GAN) for multiple mapping experiments. The generated images present recognizable and plausible images of the urban fabric. The results show that the machine can learn the inherent laws of complex urban fabrics, which expands a new applied method for the study of urban morphology.
Understanding the importance of data is crucial for realizing the full potential of AI in architectural design. Satellite images are extremely numerous, continuous, high resolution, and accessible, allowing nuanced experimentation through dataset curation. Combining deep learning with remote-sensing technologies, this study poses the following questions. Do newly available datasets uncover ideas about the city previously hidden because urban theory is predominantly Eurocentric? Do extensive and continuous datasets promise a more refined examination of datasets’ effects on outcomes? Generative adversarial networks can endlessly generate new designs based on a curated dataset, but architectural evaluation has been questionable. We employ quantitative and qualitative assessment metrics to investigate human collaboration with AI, producing results that contribute to understanding AI-based urban design models and the significance of dataset curation.
This chapter introduces the concept of morphometry, which is a subbranch in form studies in biology and has already actively employed AI technology, i.e., machine learning techniques, in the context of architectural and urban form analysis. The introduction is centered on how this concept as a distinct approach to form studies that can help trace latent granular patterns of urban types and fabrics and thereby change the understanding of city space. By presenting a research example of urban form analysis and synthesis using the concept and methods from morphometry, this chapter demonstrates the process of constructing urban form dataset, extracting features from the dataset, identifying urban types based on the extracted features, and evaluating the results of categorization. Based on this example, this chapter opens the potential that urban form analysis can be expanded into more complex dimensional studies by integrating formal and nonformal information due to the use of AI technologies.
Conference Paper
Full-text available
This paper reports on the analytical potential of machine learning methods for urban analysis. It documents a new method for data-driven urban analysis based on diagrammatic images describing each building in a city in relation to its immediate urban context. By statistically analyzing architectural and contextual features in this new dataset, the method can identify clusters of similar urban conditions and produce a detailed picture of a city's morphological structure. Remapping the clusters from data to 2D space, our method enables a new kind of urban plan that displays gradients of urban similarity. Taking Pittsburgh as a case study we demonstrate this method, and proposè`morphological types'' as a new category of urban analysis describing a given city's specific set of distinct morphological conditions. The paper concludes with a discussion of the implications of this method and its limitations, as well as its potentials for architecture, urban studies, and computation.
Conference Paper
Full-text available
In this paper we present a new parametric approach based on urban regulation, in order to generate digital models of building, directly within a three-dimensional geo-referenced environment. The system allows user, through implementation of specific generative procedures, to manage interactive architectural design processes. The designer is able to explore all possible infinite scenarios and the various possible design alternatives, by changing the parameters values and verifying in real time the results of the changes. This paper presents a short review of the main related works and an experiment.
Full-text available
This work is based on the concept that the structure of a city can be defined by six basic urban patterns. To enable more complex urban planning as a long-term objective I have developed a simulation method for generating these basic patterns and for combining them to form various structures. The generative process starts with the two-dimensional organisation of streets followed by the parceling of the remaining areas. An agent-based diffusion-contact model is the basis of these first two steps. Then, with the help of cellular automata, the sites for building on are defined and a three-dimensional building structure is derived. I illustrate the proposed method by showing how it can be applied to generate possible structures for an urban area in the city of Munich. Keywords:
This is a book about cities or, more precisely, about the physical form of cities. It starts presenting the main elements of urban form – streets, urban blocks, plots and buildings – structuring our cities and the fundamental actors and processes of transformation shaping these elements. It then applies this analytical framework to describe the evolution of cities over history as well as to explain the functioning of contemporary cities. After the initial focus on the ‘object’ (cities) the book describes how different researchers and different schools of thought have been dealing with this object since the emergence of Urban Morphology, as the science of urban form, in the turning to the twentieth century. Finally, the book tries to identify what are the most important (and specific) contributions that Urban Morphology has to offer to contemporary cities, societies and economies.
  • I Gulrajani
  • F Ahmed
  • M Arjovsky
  • V Dumoulin
  • A C Courville
Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V. and Courville, A.C.: 2017, Improved Training of Wasserstein GANs, Advances in Neural Information Processing Systems 30, 5767-5777.
  • D P Kingma
  • J Ba
Kingma, D.P. and Ba, J.: 2017, Adam: A Method for Stochastic Optimization, ArXiv:1412.6980.
Drawing with Bots: Human-computer Collaborative Drawing Experiments
  • H Zheng
Zheng, H.: 2018, Drawing with Bots: Human-computer Collaborative Drawing Experiments, Proceedings of CAADRIA 2018.