cmUNet learns modality-agnostic representations via cross-modality transformation and in-modality reconstruction with adversarial loss, enabling MarrNet to achieve superior cross-modality matching on spectrum, re-identification, and face recognition tasks.
Syncretic modality collabora- tive learning for visible infrared person re-identification,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Mind the Gap: Learning Modality-Agnostic Representations with a Cross-Modality UNet
cmUNet learns modality-agnostic representations via cross-modality transformation and in-modality reconstruction with adversarial loss, enabling MarrNet to achieve superior cross-modality matching on spectrum, re-identification, and face recognition tasks.