ConvVitMamba integrates multiscale convolution, transformer encoding, and Mamba-based refinement with PCA to outperform prior CNN, ViT, and Mamba methods in accuracy, size, and speed on four HSI benchmark datasets.
2019 2nd International Conference on Innovation in Engineering and Technology (ICIET) , pages=
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
-
ConvVitMamba: Efficient Multiscale Convolution, Transformer, and Mamba-Based Sequence modelling for Hyperspectral Image Classification
ConvVitMamba integrates multiscale convolution, transformer encoding, and Mamba-based refinement with PCA to outperform prior CNN, ViT, and Mamba methods in accuracy, size, and speed on four HSI benchmark datasets.