PINs: Progressive Implicit Networks for Multi-Scale Neural Representations

Zoe Landgraf, Alexander Sorkine Hornung, Ricardo S Cabral
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:11969-11984, 2022.

Abstract

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.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-landgraf22a, title = {{PIN}s: Progressive Implicit Networks for Multi-Scale Neural Representations}, author = {Landgraf, Zoe and Hornung, Alexander Sorkine and Cabral, Ricardo S}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {11969--11984}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/landgraf22a/landgraf22a.pdf}, url = {https://proceedings.mlr.press/v162/landgraf22a.html}, abstract = {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.} }
Endnote
%0 Conference Paper %T PINs: Progressive Implicit Networks for Multi-Scale Neural Representations %A Zoe Landgraf %A Alexander Sorkine Hornung %A Ricardo S Cabral %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-landgraf22a %I PMLR %P 11969--11984 %U https://proceedings.mlr.press/v162/landgraf22a.html %V 162 %X 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.
APA
Landgraf, Z., Hornung, A.S. & Cabral, R.S.. (2022). PINs: Progressive Implicit Networks for Multi-Scale Neural Representations. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:11969-11984 Available from https://proceedings.mlr.press/v162/landgraf22a.html.

Related Material