You are here

Davide Boscaini

Researcher
Short bio

Davide Boscaini was born in Verona (Italy) in 1988. He received a B.S. degree in Applied Mathematics in 2010 and a M.S. degree in Mathematics in 2012, both from the University of Verona (Italy). His master thesis on "Spectral Methods for Shape Analysis" was developed during an intership at the Vision, Image, Processing and Sound Lab on topics related with the analysis of 3D deformable objects. The desire to deepen his knowledge in this topic led Davide to start a Ph.D. at the University of Lugano (Switzerland) under the supervision of prof. Michael M. Bronstein. During the Ph.D. his research focussed on merging spectral approaches and deep learning techniques to develop novel algorithms for 3D Shape Analysis. Davide's efforts contributed to the birth of a new research field, called Geometric Deep Learning, aiming at extending deep learning techniques to geometric domains such as 3D shapes and graphs. He received his Ph.D. degree in Computer Science in 2017 with a dissertation on "Geometric Deep Learning for Shape Analysis". He is currently a researcher at the Technologies of Vision research unit of the Fondazione Bruno Kessler in Trento (Italy).

Research interests
shape analysis geometric deep learning
Related projects
Publications
  1. Monti, Federico; Boscaini, Davide; Masci, Jonathan; Rodolà, Emanuele; Svoboda, Jan; Bronstein, Michael M.,
    Geometric deep learning on graphs and manifolds using mixture model CNNs,
    IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
    2017
  2. Boscaini, Davide; Masci, Jonathan; Rodolà, Emanuele; Bronstein, Michael M.; Cremers, Daniel,
    Anisotropic diffusion descriptors,
    in «COMPUTER GRAPHICS FORUM»,
    vol. 35,
    n. 2,
    2016
    , pp. 431 -
    441
  3. Boscaini, Davide; Masci, Jonathan; Rodolà, Emanuele; Bronstein, Michael M.,
    Learning shape correspondence with anisotropic convolutional neural networks,
    Advances in Neural Information Processing Systems (NIPS),
    2016
    , pp. 3189-
    3197
  4. Masci, Jonathan; Rodolà, Emanuele; Boscaini, Davide; Bronstein, Michael M.; Li, Hao,
    Geometric deep learning,
    SIGGRAPH ASIA Courses,
    2016
  5. Boscaini, Davide; Eynard, Davide; Kourounis, Drosos; Bronstein, Michael M.,
    Shape-from-Operator: recovering shapes from intrinsic operators,
    in «COMPUTER GRAPHICS FORUM»,
    vol. 34,
    n. 2,
    2015
    , pp. 265 -
    274
  6. Boscaini, Davide; Masci, Jonathan; Melzi, Simone; Bronstein, Michael M.; Castellani, Umberto; Vandergheynst, Pierre,
    Learning class-specific descriptors for deformable shapes using localized spectral convolutional networks,
    in «COMPUTER GRAPHICS FORUM»,
    vol. 34,
    n. 5,
    2015
    , pp. 13 -
    23
  7. Masci, Jonathan; Boscaini, Davide; Bronstein, Michael M.; Vandergheynst, Pierre,
    Geodesic convolutional neural networks on Riemannian manifolds,
    IEEE International Conference on Computer Vision Workshop (ICCVW),
    2015
    , pp. 832-
    840
  8. Masci, Jonathan; Boscaini, Davide; Bronstein, Michael M.; Vandergheynst, Pierre,
    2015
  9. Boscaini, Davide; Castellani, Umberto,
    A sparse coding approach for local-to-global 3D shape description,
    in «THE VISUAL COMPUTER»,
    vol. 30,
    n. 11,
    2014
    , pp. 1233 -
    1245
  10. Boscaini, Davide; Girdziusas, Ramunas; Bronstein, Michael M.,
    Coulomb shapes: using electrostatic forces for deformation-invariant shape representation,
    Eurographics Workshop on 3D Object Retrieval (3DOR),
    2014
    , pp. 9-
    15

Pages