A convolutional UNet without long-range sequence modeling beat Transformer and Mamba models on a new 204-patient colorectal tumor CT dataset.
Our results demonstrate that efficient channel mixing and spatially gated features can outperform many existing computationally intensive long- range modeling techniques
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Is Long Range Sequential Modeling Necessary For Colorectal Tumor Segmentation?
A convolutional UNet without long-range sequence modeling beat Transformer and Mamba models on a new 204-patient colorectal tumor CT dataset.