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...
Үндсэн зохиолчид: | Zhang, F, Torr, P, Ranftl, R, Richter, S |
---|---|
Формат: | Conference item |
Хэл сонгох: | English |
Хэвлэсэн: |
Neural Information Processing Systems Foundation
2021
|
Ижил төстэй зүйлс
Ижил төстэй зүйлс
-
Open vocabulary semantic segmentation with Patch Aligned Contrastive Learning
-н: Mukhoti, J, зэрэг
Хэвлэсэн: (2023) -
Scalable cascade inference for semantic image segmentation
-н: Sturgess, P, зэрэг
Хэвлэсэн: (2012) -
Dense semantic image segmentation with objects and attributes
-н: Zheng, S, зэрэг
Хэвлэсэн: (2014) -
Pyramid Context Contrast for Semantic Segmentation
-н: Yuzhong Chen, зэрэг
Хэвлэсэн: (2019-01-01) -
Highly Contrast Image Correction for Dim Boundary Separation of Image Semantic Segmentation
-н: Jinyeob Choi, зэрэг
Хэвлэсэн: (2021-01-01)