Looking beyond single images for contrastive semantic segmentation learning
We present an approach to contrastive representation learning for semantic segmentation. Our approach leverages the representational power of existing feature extractors to find corresponding regions across images. These cross-image correspondences are used as auxiliary labels to guide the pixel-lev...
Κύριοι συγγραφείς: | Zhang, F, Torr, P, Ranftl, R, Richter, S |
---|---|
Μορφή: | Conference item |
Γλώσσα: | English |
Έκδοση: |
Neural Information Processing Systems Foundation
2021
|
Παρόμοια τεκμήρια
Παρόμοια τεκμήρια
-
Open vocabulary semantic segmentation with Patch Aligned Contrastive Learning
ανά: Mukhoti, J, κ.ά.
Έκδοση: (2023) -
Scalable cascade inference for semantic image segmentation
ανά: Sturgess, P, κ.ά.
Έκδοση: (2012) -
Dense semantic image segmentation with objects and attributes
ανά: Zheng, S, κ.ά.
Έκδοση: (2014) -
Pyramid Context Contrast for Semantic Segmentation
ανά: Yuzhong Chen, κ.ά.
Έκδοση: (2019-01-01) -
Highly Contrast Image Correction for Dim Boundary Separation of Image Semantic Segmentation
ανά: Jinyeob Choi, κ.ά.
Έκδοση: (2021-01-01)