Pourya Shirazian - Academia.edu (original) (raw)

Uploads

Papers by Pourya Shirazian

Research paper thumbnail of SIMD and GPU-Accelerated Rendering of Implicit Models

Implicit models inherently support automatic blending and trivial collision detection which makes... more Implicit models inherently support automatic blending and trivial collision detection which makes them an effective tool for designing complex organic shapes with many applications in various areas of research including surgical simulation systems. However, slow rendering speeds can adversely affect the performance of simulation and modelling systems. In addition, when the models are incorporated in a surgical simulation system, interactive and smooth cutting becomes a required feature for many procedures. In this research, we propose a comprehensive framework for high-performance rendering and physically-based animation of tissues modelled using implicit surfaces. Our goal is to address performance and scalability issues that arise in rendering complex implicit models as well as in dynamic interactions between surgical tool and models. Complex models can be created with implicit primitives, blending operators, affine transformations, deformations and constructive solid geometry in ...

Research paper thumbnail of Parsip: An Implicit Surfaces Polygonizer for Multi-Core Architectures

Figure 1: Building model used to profile performance of Parsip. Left figure is shaded surface wit... more Figure 1: Building model used to profile performance of Parsip. Left figure is shaded surface with overlaid mesh. Right figure shows mesh and internal skeletal control points. ABSTRACT Implicit surfaces are a mathematical representation of objects in computer graphics which are able to create smooth surfaces that can blend with each other automatically. They provide the foun-dation for models that can be used in games, animation, scientific simulation and visualization. Interactive design systems that use polygonization methods for visualization of implicit surfaces have difficulty in matching up interactive frame rates required for fast response to user changes. This is one of the reasons that implicit surfaces are not widely adapted by design artists. In this work we propose a multi-core accelerated algorithm for high performance rendering of skeletal implicit primitives that combine blend, con-structive solid geometry, warp and affine transformations in a tree data structure know...

Research paper thumbnail of EG Talk02

Research paper thumbnail of A data-efficient collaborative modelling method using websockets and the BlobTree for over-the air networks

Proceedings of the 18th International Conference on 3D Web Technology - Web3D '13, 2013

ABSTRACT Collaborative modelling has become more important in the last few years, especially now ... more ABSTRACT Collaborative modelling has become more important in the last few years, especially now that mobile devices show processing power to support 3D modelling in real-time. Current mobile networks, such as 3G and LTE, unfortunately are not as fast as traditional wired internet and have higher latency. The problem with collaborative modelling using triangle meshes is that complex models are slow to synchronize and require large network resources depending on the amount of data needed to update a model. Synchronizing thousands of triangles over the network between all participating users can introduce substantial lag between the transactions, especially on over-the air networks, making fine grained and rapid updates at interactive rates hard to achieve. In contrast the BlobTree is based on combining skeletal primitives and sketched-shapes using standard CSG and various blending operators. Using this methodology complex models can be encoded with a smaller memory footprint than mesh based systems, thus allowing for less traffic across a network to synchronize two or more workstations with one model. As a result fine grained and rapid updates are possible, improving the visual communication between all participating users.

Research paper thumbnail of Polygonization of implicit surfaces on Multi-Core Architectures with SIMD instructions

ABSTRACT In this research we tackle the problem of rendering complex models which are created usi... more ABSTRACT In this research we tackle the problem of rendering complex models which are created using implicit primitives, blending operators, affine transformations and constructive solid geometry in a design environment that organizes all these in a scene graph data structure called BlobTree. We propose a fast, scalable, parallel polygonization al- gorithm for BlobTrees that takes advantage of multicore processors and SIMD optimization techniques available on modern architectures. Efficiency is achieved through the usage of spatial data structures and SIMD optimiza- tions for BlobTree traversals and the computation of mesh vertices and other attributes. Our solution delivers interactive visualization for modeling systems based on BlobTree scene graph.

Research paper thumbnail of Sketch-based Collaborative Interactive Implicit Modelling at a Distance

With the increased complexity in state of the art models created using common digital content cre... more With the increased complexity in state of the art models created using common digital content creation applications, such as AutoCAD, Maya or XSI, the need to have more than one person work on a single model is common. Ideally people would work on the same model at the same time, from several workstations possibly at distant locations. Using a mesh based modelling approach requires synchronizing thousands of triangles over the network between all participating workstations.

Research paper thumbnail of HyperFlow: A Heterogeneous Dataflow Architecture

abstract={We propose a dataflow architecture, called HyperFlow, that offers a supporting infrastr... more abstract={We propose a dataflow architecture, called HyperFlow, that offers a supporting infrastructure that creates an abstraction layer over computation resources and naturally exposes heterogeneous computation to dataflow processing. In order to show the efficiency of our system as well as testing it, we have included a set of synthetic and real-case applications. First, we designed a general suite of micro-benchmarks that captures main parallel pipeline structures and allows evaluation of HyperFlow under different stress ...

Research paper thumbnail of SIMD and GPU-Accelerated Rendering of Implicit Models

Implicit models inherently support automatic blending and trivial collision detection which makes... more Implicit models inherently support automatic blending and trivial collision detection which makes them an effective tool for designing complex organic shapes with many applications in various areas of research including surgical simulation systems. However, slow rendering speeds can adversely affect the performance of simulation and modelling systems. In addition, when the models are incorporated in a surgical simulation system, interactive and smooth cutting becomes a required feature for many procedures. In this research, we propose a comprehensive framework for high-performance rendering and physically-based animation of tissues modelled using implicit surfaces. Our goal is to address performance and scalability issues that arise in rendering complex implicit models as well as in dynamic interactions between surgical tool and models. Complex models can be created with implicit primitives, blending operators, affine transformations, deformations and constructive solid geometry in ...

Research paper thumbnail of Parsip: An Implicit Surfaces Polygonizer for Multi-Core Architectures

Figure 1: Building model used to profile performance of Parsip. Left figure is shaded surface wit... more Figure 1: Building model used to profile performance of Parsip. Left figure is shaded surface with overlaid mesh. Right figure shows mesh and internal skeletal control points. ABSTRACT Implicit surfaces are a mathematical representation of objects in computer graphics which are able to create smooth surfaces that can blend with each other automatically. They provide the foun-dation for models that can be used in games, animation, scientific simulation and visualization. Interactive design systems that use polygonization methods for visualization of implicit surfaces have difficulty in matching up interactive frame rates required for fast response to user changes. This is one of the reasons that implicit surfaces are not widely adapted by design artists. In this work we propose a multi-core accelerated algorithm for high performance rendering of skeletal implicit primitives that combine blend, con-structive solid geometry, warp and affine transformations in a tree data structure know...

Research paper thumbnail of EG Talk02

Research paper thumbnail of A data-efficient collaborative modelling method using websockets and the BlobTree for over-the air networks

Proceedings of the 18th International Conference on 3D Web Technology - Web3D '13, 2013

ABSTRACT Collaborative modelling has become more important in the last few years, especially now ... more ABSTRACT Collaborative modelling has become more important in the last few years, especially now that mobile devices show processing power to support 3D modelling in real-time. Current mobile networks, such as 3G and LTE, unfortunately are not as fast as traditional wired internet and have higher latency. The problem with collaborative modelling using triangle meshes is that complex models are slow to synchronize and require large network resources depending on the amount of data needed to update a model. Synchronizing thousands of triangles over the network between all participating users can introduce substantial lag between the transactions, especially on over-the air networks, making fine grained and rapid updates at interactive rates hard to achieve. In contrast the BlobTree is based on combining skeletal primitives and sketched-shapes using standard CSG and various blending operators. Using this methodology complex models can be encoded with a smaller memory footprint than mesh based systems, thus allowing for less traffic across a network to synchronize two or more workstations with one model. As a result fine grained and rapid updates are possible, improving the visual communication between all participating users.

Research paper thumbnail of Polygonization of implicit surfaces on Multi-Core Architectures with SIMD instructions

ABSTRACT In this research we tackle the problem of rendering complex models which are created usi... more ABSTRACT In this research we tackle the problem of rendering complex models which are created using implicit primitives, blending operators, affine transformations and constructive solid geometry in a design environment that organizes all these in a scene graph data structure called BlobTree. We propose a fast, scalable, parallel polygonization al- gorithm for BlobTrees that takes advantage of multicore processors and SIMD optimization techniques available on modern architectures. Efficiency is achieved through the usage of spatial data structures and SIMD optimiza- tions for BlobTree traversals and the computation of mesh vertices and other attributes. Our solution delivers interactive visualization for modeling systems based on BlobTree scene graph.

Research paper thumbnail of Sketch-based Collaborative Interactive Implicit Modelling at a Distance

With the increased complexity in state of the art models created using common digital content cre... more With the increased complexity in state of the art models created using common digital content creation applications, such as AutoCAD, Maya or XSI, the need to have more than one person work on a single model is common. Ideally people would work on the same model at the same time, from several workstations possibly at distant locations. Using a mesh based modelling approach requires synchronizing thousands of triangles over the network between all participating workstations.

Research paper thumbnail of HyperFlow: A Heterogeneous Dataflow Architecture

abstract={We propose a dataflow architecture, called HyperFlow, that offers a supporting infrastr... more abstract={We propose a dataflow architecture, called HyperFlow, that offers a supporting infrastructure that creates an abstraction layer over computation resources and naturally exposes heterogeneous computation to dataflow processing. In order to show the efficiency of our system as well as testing it, we have included a set of synthetic and real-case applications. First, we designed a general suite of micro-benchmarks that captures main parallel pipeline structures and allows evaluation of HyperFlow under different stress ...