Transcription of M -S C A D C - arXiv
{{id}} {{{paragraph}}}
Example: tourism industry
M -S C A D C - arXiv
Published as a conference paper at ICLR 2016. M ULTI -S CALE C ONTEXT AGGREGATION BY. D ILATED C ONVOLUTIONS. Fisher Yu Princeton University Vladlen Koltun Intel Labs [ ] 30 Apr 2016. A BSTRACT. State-of-the-art models for semantic segmentation are based on adaptations of convolutional networks that had originally been designed for image classifica- tion. However, dense prediction problems such as semantic segmentation are structurally different from image classification. In this work, we develop a new convolutional network module that is specifically designed for dense prediction. The presented module uses dilated convolutions to systematically aggregate multi- scale contextual information without losing resolution .
scale contextual reasoning in combination with full-resolution output. Recent work has studied two approaches to dealing with the conflicting demands of multi-scale reasoning and full-resolution dense prediction. One approach involves repeated up …
Loading..
Tags:
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document: