DSAFormer applies spatial and channel sparse multi-head cross-attention plus a learnable fusion block to reduce redundant token interactions and improve multispectral object detection on four public datasets.
Title resolution pending
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
-
Dual Sparse Aggregation Transformer for Multispectral Object Detection
DSAFormer applies spatial and channel sparse multi-head cross-attention plus a learnable fusion block to reduce redundant token interactions and improve multispectral object detection on four public datasets.