Abstract
We propose a monocular depth estimation algorithm based on whole strip masking (WSM) and reliability-based refinement. First, we develop a convolutional neural network (CNN) tailored for the depth estimation. Specifically, we design a novel filter, called WSM, to exploit the tendency that a scene has similar depths in horizonal or vertical directions. The proposed CNN combines WSM upsampling blocks with a ResNet encoder. Second, we measure the reliability of an estimated depth, by appending additional layers to the main CNN. Using the reliability information, we perform conditional random field (CRF) optimization to refine the estimated depth map. Experimental results demonstrate that the proposed algorithm provides the state-of-the-art depth estimation performance.
| Original language | English |
|---|---|
| Title of host publication | Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings |
| Editors | Vittorio Ferrari, Cristian Sminchisescu, Yair Weiss, Martial Hebert |
| Publisher | Springer Verlag |
| Pages | 39-55 |
| Number of pages | 17 |
| ISBN (Print) | 9783030012243 |
| DOIs | |
| Publication status | Published - 2018 Jan 1 |
| Event | 15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany Duration: 2018 Sep 8 → 2018 Sep 14 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 11208 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Other
| Other | 15th European Conference on Computer Vision, ECCV 2018 |
|---|---|
| Country | Germany |
| City | Munich |
| Period | 18/9/8 → 18/9/14 |
Fingerprint
Keywords
- Depth map refinement
- Monocular depth estimation
- Reliability
- Whole strip masking
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)
Cite this
Monocular Depth Estimation Using Whole Strip Masking and Reliability-Based Refinement. / Heo, Minhyeok; Lee, Jaehan; Kim, Kyung Rae; Kim, Han Ul; Kim, Chang-Su.
Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. ed. / Vittorio Ferrari; Cristian Sminchisescu; Yair Weiss; Martial Hebert. Springer Verlag, 2018. p. 39-55 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11208 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Monocular Depth Estimation Using Whole Strip Masking and Reliability-Based Refinement
AU - Heo, Minhyeok
AU - Lee, Jaehan
AU - Kim, Kyung Rae
AU - Kim, Han Ul
AU - Kim, Chang-Su
PY - 2018/1/1
Y1 - 2018/1/1
N2 - We propose a monocular depth estimation algorithm based on whole strip masking (WSM) and reliability-based refinement. First, we develop a convolutional neural network (CNN) tailored for the depth estimation. Specifically, we design a novel filter, called WSM, to exploit the tendency that a scene has similar depths in horizonal or vertical directions. The proposed CNN combines WSM upsampling blocks with a ResNet encoder. Second, we measure the reliability of an estimated depth, by appending additional layers to the main CNN. Using the reliability information, we perform conditional random field (CRF) optimization to refine the estimated depth map. Experimental results demonstrate that the proposed algorithm provides the state-of-the-art depth estimation performance.
AB - We propose a monocular depth estimation algorithm based on whole strip masking (WSM) and reliability-based refinement. First, we develop a convolutional neural network (CNN) tailored for the depth estimation. Specifically, we design a novel filter, called WSM, to exploit the tendency that a scene has similar depths in horizonal or vertical directions. The proposed CNN combines WSM upsampling blocks with a ResNet encoder. Second, we measure the reliability of an estimated depth, by appending additional layers to the main CNN. Using the reliability information, we perform conditional random field (CRF) optimization to refine the estimated depth map. Experimental results demonstrate that the proposed algorithm provides the state-of-the-art depth estimation performance.
KW - Depth map refinement
KW - Monocular depth estimation
KW - Reliability
KW - Whole strip masking
UR - http://www.scopus.com/inward/record.url?scp=85055438092&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85055438092&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-01225-0_3
DO - 10.1007/978-3-030-01225-0_3
M3 - Conference contribution
SN - 9783030012243
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 39
EP - 55
BT - Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
A2 - Ferrari, Vittorio
A2 - Sminchisescu, Cristian
A2 - Weiss, Yair
A2 - Hebert, Martial
PB - Springer Verlag
ER -