Scipediacontent: Scipediacontent moved page Draft Content 293480827 to Liebel Korner 2019a

2021-01-28T18:21:17Z

Scipediacontent moved page Draft Content 293480827 to Liebel Korner 2019a

Scipediacontent: Created page with " == Abstract == We introduce MultiDepth, a novel training strategy and convolutional neural network (CNN) architecture that allows approaching single-image depth estimation (..."

2021-01-28T18:21:12Z

Created page with " == Abstract == We introduce MultiDepth, a novel training strategy and convolutional neural network (CNN) architecture that allows approaching single-image depth estimation (..."

New page

== Abstract ==

We introduce MultiDepth, a novel training strategy and convolutional neural network (CNN) architecture that allows approaching single-image depth estimation (SIDE) as a multi-task problem. SIDE is an important part of road scene understanding. It, thus, plays a vital role in advanced driver assistance systems and autonomous vehicles. Best results for the SIDE task so far have been achieved using deep CNNs. However, optimization of regression problems, such as estimating depth, is still a challenging task. For the related tasks of image classification and semantic segmentation, numerous CNN-based methods with robust training behavior have been proposed. Hence, in order to overcome the notorious instability and slow convergence of depth value regression during training, MultiDepth makes use of depth interval classification as an auxiliary task. The auxiliary task can be disabled at test-time to predict continuous depth values using the main regression branch more efficiently. We applied MultiDepth to road scenes and present results on the KITTI depth prediction dataset. In experiments, we were able to show that end-to-end multi-task learning with both, regression and classification, is able to considerably improve training and yield more accurate results.

Comment: Accepted for presentation at the IEEE Intelligent Transportation Systems Conference (ITSC) 2019

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1907.11111 http://arxiv.org/abs/1907.11111]

* [http://arxiv.org/pdf/1907.11111 http://arxiv.org/pdf/1907.11111]

* [http://xplorestaging.ieee.org/ielx7/8907344/8916833/08917177.pdf?arnumber=8917177 http://xplorestaging.ieee.org/ielx7/8907344/8916833/08917177.pdf?arnumber=8917177],
: [http://dx.doi.org/10.1109/itsc.2019.8917177 http://dx.doi.org/10.1109/itsc.2019.8917177]

* [https://dblp.uni-trier.de/db/journals/corr/corr1907.html#abs-1907-11111 https://dblp.uni-trier.de/db/journals/corr/corr1907.html#abs-1907-11111],
: [https://academic.microsoft.com/#/detail/2989929904 https://academic.microsoft.com/#/detail/2989929904]

← Older revision	Revision as of 18:21, 28 January 2021
(No difference)

Liebel Korner 2019a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 293480827 to Liebel Korner 2019a

Scipediacontent: Created page with " == Abstract == We introduce MultiDepth, a novel training strategy and convolutional neural network (CNN) architecture that allows approaching single-image depth estimation (..."