Skip to yearly menu bar Skip to main content


PINs: Progressive Implicit Networks for Multi-Scale Neural Representations

Zoe Landgraf · Alexander Sorkine Hornung · ricardo cabral

Hall E #131

Keywords: [ MISC: Representation Learning ] [ DL: Other Representation Learning ] [ MISC: Supervised Learning ] [ APP: Computer Vision ]


Multi-layer perceptrons (MLP) have proven to be effective scene encoders when combined with higher-dimensional projections of the input, commonly referred to as positional encoding.However, scenes with a wide frequency spectrum remain a challenge: choosing high frequencies for positional encoding introduces noise in low structure areas, while low frequencies results in poor fitting of detailed regions. To address this, we propose a progressive positional encoding, exposing a hierarchical MLP structure to incremental sets of frequency encodings.Our model accurately reconstructs scenes with wide frequency bands and learns a scene representation at progressive level of detail without explicit per-level supervision. The architecture is modular: each level encodes a continuous implicit representation that can be leveraged separately for its respective resolution, meaning a smaller network for coarser reconstructions.Experiments on several 2D and 3D datasets shows improvements in reconstruction accuracy, representational capacity and training speed compared to baselines.

Chat is not available.