• Noise

    The paper Volumetric Spot Noise for procedural 3D Shell Texture Synthesis has been accepted at Computer CGVC16.

    we present an extension of the Locally Controlled Spot Noise and a visualization pipeline for volumetric fuzzy details synthesis. We extend the noise model to author volumetric fuzzy details using filtered 3D quadratic kernel functions convolved with a projective non-uniform 2D distribution of impulses. We propose a new method based on order independent splatting to compute a fast view dependent approximation of shell noise at interactive rates. Our method outperforms ray marching techniques and avoids aliasing artifacts, thus improving interactive content authoring feedback. Generated surface details share the same properties as procedural noise: they extend on potentially infinite surfaces, are defined in an extremely compact way, are non-repetitive, continuous and independent of the definition of the underlying surface.

  • ANR Huge Digital Worlds

    The ANR project Huge Digital Worlds has been accepted and funded by the Agence Nationale pour la Recherche.

    In this project, we address the generation, processing and rendering of huge realistic landscapes using an original approach: program-based generative content production, also called procedural modeling. Instead relying on explicit representations, procedural and implicit models generate geometry and/or texture on the fly by using functions and procedures controled by a reduced set of parameters.

  • Modeling entangled details

    Our paper Efficient modeling of entangled details for natural scenes has been accepted at Computer Graphics Forum.

    Digital landscape realism often comes from the multitude of details that are hard to model such as fallen leaves, rock piles or entangled fallen branches. In this article, we present a method for augmenting natural scenes with a huge amount of details such as grass tufts, stones, leaves or twigs. Our approach takes advantage of the observation that those details can be approximated by replications of a few similar objects and therefore relies on mass-instancing. We propose an original structure, the Ghost Tile, that stores a huge number of overlapping candidate objects in a tile, along with a pre-computed collision graph. Details are created by traversing the scene with the Ghost Tile and generating instances according to user-defined density fields that allow to sculpt layers and piles of entangled objects while providing control over their density and distribution.

  • Two articles accepted at Eurographics 2016

    Two papers, Sparse representation of terrains for procedural modeling and Large Scale Terrain Generation from Tectonic Uplift and Fluvial Erosion have been accepted at Eurographics.

    In the sparse modeling paper, we present a simple and efficient method to represent terrains as elevation functions built from linear combinationsof landform features (atoms). These features can be extracted either from real world data-sets or procedural primitives, orfrom any combination of multiple terrain models. Our approach consists in representing the elevation function as a sparsecombination of primitives, a concept which we call Sparse Construction Tree, which blends the different landform featuresstored in a dictionary. The sparse representation allows us to represent complex terrains using combinations of atoms from asmall dictionary, yielding a powerful and compact terrain representation and synthesis tool. Moreover, we present a methodfor automatically learning the dictionary and generating the Sparse Construction Tree model. We demonstrate the efficiency ofour method in several applications: inverse procedural modeling of terrains, terrain amplification and synthesis from a coarsesketch.

  • Environment Sensitive Objects

    Our paper Environmental Objects for Authoring Procedural Scene has been accepted at Computer Graphics Forum.

    We propose a novel approach for authoring large scenes with automatic enhancement of objects to create geometric decoration details such as snow cover, icicles, fallen leaves, grass tufts, or even trash. We introduce environmental objects that extend an input object geometry with a set of procedural effects that defines how the object reacts to the environment, and by a set of scalar fields that defines the influence of the object over of the environment. The user controls the scene by modifying environmental variables, such as temperature or humidity fields. The scene definition is hierarchical: objects can be grouped and their behaviors can be set at each level of the hierarchy. Our per object definition allows us to optimize and accelerate the effects computation, which also enables us to generate large scenes with many geometric details at a very high level of detail. In our implementation, a complex urban scene of ten-thousand square meters, represented with details of less than one centimeter, can be locally modified and entirely re-generated in a few seconds.

  • Invited speaker at UBISOFT Procedural Days

    I made a presentation on procedural generation at UBISOFT Paris: Modeling and generating scenic landscapes.

    In this talk, I will present our latest research on procedural generation of terrains, roads and rivers. I will show that some phenomenological approaches lend themselves for interactive and intuitive control and can be used to generate complex models.


    The project PAPAYA has been accepted.

  • Terrain Modeling from Feature Primitives

    Our paper has been accepted at Computer Graphics Forum.

    We introduce a compact hierarchical procedural model that combines feature-based primitives to describe complex terrains with varying level of detail. Our model is inspired by skeletal implicit surfaces and defines the terrain elevation function by using a construction tree. Leaves represent terrain features and they are generic parameterized skeletal primitives such as mountains, ridges, valleys, rivers, lakes, or roads. Inner nodes combine the leaves and sub-trees by carving, blending, or warping operators. The elevation of the terrain at a given point is evaluated by traversing the tree and by combining the contributions of the primitives. The definition of the tree leaves and operators guarantees that the resulting elevation function follows the Lipschitz property which speeds up sphere tracing and adaptive tessellation algorithms used to render the terrain. Our model is compact and allows for the creation of large terrains with a high level of detail using a reduced set of primitives. We show the creation of different kinds of landscapes and demonstrate that our model allows to efficiently control the shape and distribution of landform features.

  • Invited speaker at the Collège de France

    I will give a talk at the prestigious Collège de France on procedural generation of virtual worlds.

    Watch the video of the presentation.

    Marie Paule Cani invited me to give a presentation at the Collège de France. In this presentation, I present an overview of challenges in the generation of virtual world and show how procedural modeling can provide some efficient solutions to various complex problems. The slides and vidéo of the presentation and details can be found here.