The Devil is in the Decoder

Zbigniew Wojna, Jasper Uijlings, Sergio Guadarrama, Nathan Silberman, Liang-Chieh Chen, Alireza Fathi and Vittorio Ferrari

Abstract

Many machine vision applications require predictions for every pixel of the input image (for example semantic segmentation, boundary detection). Models for such problems usually consist of encoders which decreases spatial resolution while learning a high-dimensional representation, followed by decoders who recover the original input resolution and result in low-dimensional predictions. While encoders have been studied rigorously, relatively few studies address the decoder side. Therefore this paper presents an extensive comparison of a variety of decoders for a variety of pixel-wise prediction tasks. Our contributions are: (1) Decoders matter: we observe significant variance in results between different types of decoders on various problems. (2) We introduce a novel decoder: bilinear additive upsampling. (3) We introduce new residual-like connections for decoders.

Session

Spotlights

Files

Paper (PDF)
Supplementary (PDF)
Video (MP4)

DOI

10.5244/C.31.10
https://dx.doi.org/10.5244/C.31.10

Citation

Bibtex

            @inproceedings{BMVC2017_10,
                title={The Devil is in the Decoder},
                author={Zbigniew Wojna, Jasper Uijlings, Sergio Guadarrama, Nathan Silberman, Liang-Chieh Chen, Alireza Fathi and Vittorio Ferrari},
                year={2017},
                month={September},
                pages={10.1-10.13},
                articleno={10},
                numpages={13},
                booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
                publisher={BMVA Press},
                editor={Tae-Kyun Kim, Stefanos Zafeiriou, Gabriel Brostow and Krystian Mikolajczyk},
                doi={10.5244/C.31.10},
                isbn={1-901725-60-X},
                url={https://dx.doi.org/10.5244/C.31.10}
            }