MPAE discovers object parts by filling randomly masked image patches with learned part descriptors and then restoring the image, reporting state-of-the-art unsupervised part discovery on four benchmarks.
Deep vit features as dense visual descriptors
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
contest 1representative citing papers
citing papers explorer
-
Unsupervised Part Discovery via Descriptor-Based Masked Image Restoration with Optimized Constraints
MPAE discovers object parts by filling randomly masked image patches with learned part descriptors and then restoring the image, reporting state-of-the-art unsupervised part discovery on four benchmarks.