Siddhartha Chaudhuri

Adobe Research and IIT Bombay, India

Bio: Siddhartha Chaudhuri is a Senior Research Scientist in the Creative Intelligence Lab at Adobe Research, and an Assistant Professor (on extended leave) of Computer Science and Engineering at IIT Bombay. He obtained his Ph.D. from Stanford University, and his undergraduate degree from IIT Kanpur. He subsequently did postdoctoral research at Stanford and Princeton, and taught at Cornell. Siddhartha's work combines geometric analysis, machine learning, and UI innovation to make sophisticated 3D geometric modeling accessible even to non-expert users. He also studies foundational problems in geometry processing (retrieval, segmentation, correspondences) that arise from this pursuit. His research themes include assembly-based modeling, semantic attributes for design, generative neural networks for shapes, and other applications of deep learning to 3D geometry processing. He is the original author of the commercial 3D character modeling tool Adobe Fuse, and the author of the open-source Thea toolkit for geometric computing.

Data-Driven 3D Modeling

Date: 15^th Nov, 08:00-08:45 am (EST)

Abstract: For this keynote, I have chosen a “broad” topic rather than a “focused” one. I will describe foundational challenges in 3D interactive modeling, and discuss why classical hand-designed techniques and interfaces do not always satisfactorily address them. I will then present ways in which the research community in general, a nd my colleagues and I in particular, have attempted to train algorithms on large repositories of data in order to better capture design principles and intent that are relevant to human modelers. Since the context is interactive modeling, I will chiefly discuss generative or quasi-generative methods. The talk will span about a decade of work in machine learning as applied to 3D modeling, outline exciting new work that is happening right now, and conjecture where the field might head next.

Lining Yao

Carnegie Mellon University, USA

Bio: Lining Yao is an Assistant Professor of Human-Computer Interaction Institute at Carnegie Mellon University (CMU), School of Computer Science, directing the Morphing Matter Lab. Lining also holds courtesy appointments at Mechanical Engineering and Material Sciences and Engineering at CMU. Morphing Matter lab develops processes, materials, tools, and applications of adaptive, dynamic, and intelligent morphing matter from nano to macro scales. Research often combines material science, computational fabrication, and creative design practices. The mission is to advance both science and society with the design of morphing matter. Lining and her lab work anti-disciplinarily, publishing and exhibiting across science, engineering, design, and art. Lining gained her Ph.D. from the MIT Media Lab. She is a Wired UK fellow, CMU Provost's Inclusive Teaching Fellow, and a recipient of the NSF CAREER Award.

Computational Geometry for Morphing Matter Design

Date: 16^th Nov, 09:45-10:45 am (EST)

Abstract: Morphing Matter is an interplay of geometry and hidden forces. Lining Yao, the director of Morphing Matter Lab, will share her team's experiences of computing, designing, and fabricating morphing mechanisms that leverage both geometrical and physical knowledge of materials. Lining will unfold a few marriages of geometries and forces in the talk: a conformal map interconnecting beams shrinking and fighting for the lowest entropy, a frustum-shaped groove interfering disks swelling with differential diffusion rate, and a triangulated filler path affecting spacer fabric deforming with biased shear forces. Novel morphing mechanisms and applications also come from these marriages, such as self-assembling furniture, crawling soft robots, and morphing pasta.

Bianca Falcidieno

National Research Council, Italy.

Bio: Bianca Falcidieno is a former Research Director at the Institute of Applied Mathematics and Information Technologies of the National Research Council of Italy (CNR-IMATI), and she was the chair of the CNR Research Area of Genova (2009-2013). Her research focuses on shape modeling, computer graphics, computational mathematics, and knowledge representation. She has firmly established Genova as an internationally leading research center in these topics, strongly interacting with industrial and social application fields: from industrial design and manufacturing to geographic information systems, from multimedia to cultural heritage. In her career, she coordinated several international projects and the CNR participation to many national research initiatives. As the coordinator of the European project AIM@SHAPE, she pioneered the development of semantics-driven approaches for representing 3D shapes, based on coupling knowledge technologies with geometric representations in this area. She was in charge of various international commitments, including editorial tasks, the organization of several international conferences and workshops as the chair or co-chair, and participation to international program committees. Bianca has also been a gifted mentor, and several of her former students and postdocs have meanwhile become internationally established researchers of their own. Together with Prof. Kunii, she was the co-founder and the editor in chief of the Shape Modeling International Journal and created the Shape Modeling International conference series. Bianca Falcidieno has authored over 250 peer-reviewed scientific publications, her work has been published extensively in the leading journals of the field. For the 80th CNR anniversary, she was included in the 12 top-level female researchers in the history of CNR. Bianca is a Eurographics Fellow, a Pioneer of the Solid Modeling association and in 2019 awarded with the Eurographics Gold Medal for her outstanding scientific contributions to computer graphics and shape modeling.

Shape Modeling: from Geometry to Semantics

Date: 16^th Nov, 08:00-08:30 am (EST)

Abstract: Geometric Modeling and Computer Graphics technologies have allowed us to represent and visualize objects and scenes on our digital devices since the 70s. However, during the 90s, the technological evolution of graphical devices and systems, the growth of applications and users, and the complexity of the models brought out the need for a shift of focus: from the modeling of geometries to the description of objects, to highlight aspects of shapes less formalized than geometry.
To meet this need, thanks to the collaboration with Professor Kunii, in the mid-90s, I was able to start a new line of research with him, Shape Modeling. Based on the concept of shape, Shape Modeling is not only related to objects; geometry but also to their structure (object features and part-whole decomposition), meaning in a specific context, and interaction with time. This research aimed to enhance traditional modeling approaches through representations that are closer to our semantic coding of shape perception in both cognitive and creative processes.
In this presentation, I will talk about the beginning and developments of Shape Modeling research and its evolution towards the representation of semantics associated with shapes, including computational topology studies for shape understanding, that I started thanks to Professor Kunii.

Mélina Skouras

Inria Grenoble Rhône-Alpes, France

Bio: Mélina Skouras is a permanent researcher in the ANIMA team at Inria Grenoble, France. Until October 2017, she was a postdoctoral associate at MIT where she worked with Prof. Wojciech Matusik on the computational design of meta-materials. She obtained her PhD in 2014 from the Computer Graphics Laboratory of ETH Zurich, Switzerland, where she collaborated with Disney Research Zurich. Her thesis focused on the development of novel algorithms for the design of custom deformable objects. She received her Master's degree in Computer Science and Applied Mathematics from ENSIMAG, INP Grenoble, France, in 2004. Before starting her PhD, she was a software developer at Dassault Systèmes, in the CATIA Geometric Modeler team.

On the computational design of deformable objects

Date: 15^th Nov, 10:00-10:45 am (EST)

Abstract: Recent technological advances in digital manufacturing enable experts and casual users alike to design and fabricate complex structures with a fine control over their external geometry and internal material distribution. However, designing an object that is truly functional remains a challenging task. This is because, in the real world, the object is subject to physical laws that affect its behavior, e.g. determine its final shape. Therefore, the designer needs to estimate and invert the effects of the physics on the object to obtain a structure that behaves as desired. This is extremely difficult to achieve by hand and explains why this task is often based on a trial-and-error iterative process. In this talk, I will explain how inverse modeling tools can be used to automate the design of custom elastic objects with prescribed target geometries. I will focus on some important challenges that arise when designing such tools and present the principles that we followed to address them when working on concrete applications.