Effectively discerning spatial-spectral dependencies in HSI denoising is
crucial, but prevailing methods using convolution or transformers still face
computational efficiency limitations. Recently, the emerging Selective State
Space Model(Mamba) has risen with its nearly linear computational complexity in
processing natural language sequences, which inspired us to explore its
potential in handling long spectral sequences. In this paper, we propose
HSIDMamba(HSDM), tailored to exploit the linear complexity for effectively
capturing spatial-spectral dependencies in HSI denoising. In particular, HSDM
comprises multiple Hyperspectral Continuous Scan Blocks, incorporating
BCSM(Bidirectional Continuous Scanning Mechanism), scale residual, and spectral
attention mechanisms to enhance the capture of long-range and local
spatial-spectral information. BCSM strengthens spatial-spectral interactions by
linking forward and backward scans and enhancing information from eight
directions through SSM, significantly enhancing the perceptual capability of
HSDM and improving denoising performance more effectively. Extensive
evaluations against HSI denoising benchmarks validate the superior performance
of HSDM, achieving state-of-the-art results in performance and surpassing the
efficiency of the latest transformer architectures by $30\%$.

通过利用选择性状态空间模型（Mamba）的近乎线性计算复杂性，本文提出了 HSIDMamba（HSDM）用于高光谱图像去噪，利用 HSDM 的连续扫描机制强化了空间 - 光谱相互作用，提高了去噪性能，实验结果表明 HSDM 在性能上超过了最新的变压器架构的效率提升 30%。

HSIDMamba：探索用于高光谱去噪的双向状态空间模型

HSIDMamba: Exploring Bidirectional State-Space Models for Hyperspectral  Denoising

In this paper, we develop a structure theory for generalized spectral
sequences, which are derived from chain complexes that are filtered over
arbitrary partially ordered sets. Also, a more general construction method
reminiscent of exact couples is studied, together with examples where they
arise naturally. As for ordinary spectral sequences we will see differentials
and group extensions, however the real power comes from the appearance of
natural isomorphims between pages of differing indices.
The constructions reveal finer invariants than ordinary spectral sequences,
and they connect to other fields such as Fary functors and perverse sheaves.
They are based on a natural index scheme, which allows us to obtain new results
even in the standard case of Z-filtered chain complexes, e.g. a useful
criterion for a product structure for Grothendieck's spectral sequences, and
news paths to connect the first or second page to the limit.
This turns out to yield the right framework for unifying several spectral
sequences that one would usually apply one after another. Examples that we work
out are successive Leray--Serre spectral sequences, the Adams--Novikov spectral
sequence following the chromatic spectral sequence, successive Grothendieck
spectral sequences, and successive Eilenberg--Moore spectral sequences.

本文发展了一种广义谱序列的结构理论，这些谱序列是从在任意局部有序集上过滤的链复合物中推导出的，并研究了一种更一般的建构方法，从而清楚地表现出其精细的不变量和其与其他领域的联系。