Simple and Efficient Vision Backbone Adapter for Image Semantic Segmentation.

Dingjie Peng Wataru Kameyama

Published in: ACML (2023)

Keyphrases

semantic segmentation
single image
input image
image features
multiscale
superpixels
scene classification
image data
image classification
image content
conditional random fields
segmentation method
image segmentation
image collections
image regions
image retrieval
image representation
test images
weakly supervised
pixel level
keypoints
image processing
image database
feature space
high resolution