Brainstorming on materials

A lot of people inside the team are working on materials from different standpoints.
We’ve thus been brainstorming on this topic to know who’s interested in what aspects of materials exactly.

Below is a non-exhaustive categorization of research directions, along with team members’ interests.
Feel free to edit it or even add categories! This is meant to foster internal collaborations.

Data bases

  • Romain has assembled a data-base of retro-reflective materials
    He has also started measuring micro-facet distributions
    Pascal has sent the same materials for GelSight capture to Ted Adelson
  • Ivo & Xavier would like to build a public ground truth database of materials for comparison
    Ivo has also measured various materials immersed in different liquids
    He would be interested to use kaleido-camera images for reflectance measurement purposes

Micro-structures

  • Romain is interested in evaluating the micro-facet theory with various data
    He’s also looking forward to connect Maxwell equations to BSSRDF models.
  • Nicolas, Romain & Pascal are interested in studying statistics of micro-heightfields
    The goal would be to summarize them in a way that corresponds to meaningful material properties
  • Laurent & Pascal are interested in how volume statistics of micro-structures give rise to BSSRDF lobes
    This is related to PhD work of Charles de Rousiers

Analysis

  • Romain & Xavier would like to analyze how well existing BRDF models represent retro lobes
  • Laurent & Xavier are interested in first-order/Fourier analysis of BRDF models
    Laurent is looking for a model which would be parametrized the same way in spatial or Fourier domains
    He believes that Fourier space could also be adapted to guide importance sampling
  • Xavier, Gael & Pascal have initiated a first-order analysis of the rendering equation
    Various types of BRDF gradients are involved, and now deserve further inquiry

Parametrization

  • Romain would like to know how the retro-reflective lobe should be parametrised
    With Christophe Schlick, they have been working on new families of BRDF parametrizations
  • Ivo & Xavier have been thinking about parametrization issues (singularities)
    • Ivo’s comment: Rusinkiewicz’s param. is worse than  Zickler’s parametrization in my opinion, the latter being the best existing one, it still has singularities though and cannot be easily used for a basis expansion
  • Carles Botsch & Pascal are studying Low’s parametrization compared to Rusinkievicz’s

Fitting

  • Romain & Laurent are working on a fitting library that includes Rational BRDFs
    They are also trying to compress color data that is redundant in measurements.
  • Ivo is working on the incorporation of polarisation in BRDFs (pBRDF)
    He’s also interested in data fitting and its accuracy in general.
  • Gael & Xavier are interested in the technical challenges posed by fitting.

Scale transfers

  • Gael & Xavier are interested in the representation and acquisition of macro-structures
    In particular, he would like to find solutions to transition from these structures to SV-BRDFs
  • Laurent & Xavier would like to understand how and when transitions from BSSRDFs to BRDFs should occur
  • Laurent & Pascal would like to find a material model that incorporates light and view footprints.
    It would make simulations a lot faster, and provide for a description of material appearance at various scales

Imaging

  • Romain Vergne,  Carlos, Xavier  & Pascal work on the reconstruction of radiance maps from rendered layers
    Once reconstructed, they can be processed to modify material properties through (spherical) image filtering
  • Adolfo Muñoz, Roland Fleming and Pascal are studying image statistics that could act as perceptual cues for materials
    The end-goal would be to connect these cues to B[SS]RDF properties via the radiance map intermediate structure.

Restitution

  • Ivo & Xavier would like to investigate ways to change material appearance by immersing objects
  • Boris, Gael, Xavier & Pascal are interested in fabricating anisotropic materials from  vector fields

 

Comments are closed.