# SDC - Stacked Dilated Convolution: A Unified Descriptor Network for Dense Matching Tasks

Source: [SDC - Stacked Dilated Convolution: A Unified Descriptor Network for Dense Matching Tasks](<https://www.sota2.com/research/paper/68fa658b1e209131cea0afd8>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T21:29:43.441Z

Authors: Ren\\'e Schuster, Oliver Wasenm\\"uller, Christian Unger, Didier Stricker

Published: 2019-04-05T14:07:02.000Z

## Abstract

Dense pixel matching is important for many computer vision tasks such as disparity and flow estimation. We present a robust, unified descriptor network that considers a large context region with high spatial variance. Our network has a very large receptive field and avoids striding layers to maintain spatial resolution. These properties are achieved by creating a novel neural network layer that consists of multiple, parallel, stacked dilated convolutions (SDC). Several of these layers are combined to form our SDC descriptor network. In our experiments, we show that our SDC features outperform state-of-the-art feature descriptors in terms of accuracy and robustness. In addition, we demonstrate the superior performance of SDC in state-of-the-art stereo matching, optical flow and scene flow algorithms on several famous public benchmarks.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Occlusion Detection on SMPL Intra-Subject (test)](<https://www.sota2.com/research/sota/occlusion-detection-on-smpl-intra-subject-test.md>): Occlusion Detection; SMPL Intra-Subject (test); Average Precision: 56.4.

- [Occlusion Detection on The Relightables Intra-Subject (test)](<https://www.sota2.com/research/sota/occlusion-detection-on-the-relightables-intra-subject-test.md>): Occlusion Detection; The Relightables Intra-Subject (test); AP: 48.17.

- [Occlusion Detection on RenderPeople Intra-Subject (test)](<https://www.sota2.com/research/sota/occlusion-detection-on-renderpeople-intra-subject-test.md>): Occlusion Detection; RenderPeople Intra-Subject (test); AP: 58.38.

- [Occlusion Detection on SMPL Inter-Subject (test)](<https://www.sota2.com/research/sota/occlusion-detection-on-smpl-inter-subject-test.md>): Occlusion Detection; SMPL Inter-Subject (test); AP: 28.98.

- [Dense Correspondence Search on SMPL Intra-Subject (test)](<https://www.sota2.com/research/sota/dense-correspondence-search-on-smpl-intra-subject-test.md>): Dense Correspondence Search; SMPL Intra-Subject (test); EPE (non-occluded): 16.96.

- [Dense Correspondence Search on Relightables Intra-Subject (test)](<https://www.sota2.com/research/sota/dense-correspondence-search-on-relightables-intra-subject-test.md>): Dense Correspondence Search; Relightables Intra-Subject (test); EPE (Non-Occluded): 17.79.

- [Dense Correspondence Search on RenderPeople Intra-Subject (test)](<https://www.sota2.com/research/sota/dense-correspondence-search-on-renderpeople-intra-subject-test.md>): Dense Correspondence Search; RenderPeople Intra-Subject (test); EPE (Non-Occluded): 20.07.

- [Dense Correspondence Search on SMPL Inter-Subject (test)](<https://www.sota2.com/research/sota/dense-correspondence-search-on-smpl-inter-subject-test.md>): Dense Correspondence Search; SMPL Inter-Subject (test); EPE (non-occluded): 81.48.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1904.03076>)

[PDF](<https://arxiv.org/pdf/1904.03076.pdf>)