Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability

Paul Hill; Zhiming Liu; Nantheera Anantrasirichai

doi:10.48550/arXiv.2503.17700

Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability

Paul Hill, Zhiming Liu, Nantheera Anantrasirichai

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

6 Downloads (Pure)

Abstract

Restoration and enhancement are essential for improving the quality of videos captured under atmospheric turbulence conditions, aiding visualization, object detection, classification, and tracking in surveillance systems. In this paper, we introduce a novel Mamba-based method, the 3D Mamba-Based Atmospheric Turbulence Removal (MAMAT), which employs a dual-module strategy to mitigate these distortions. The first module utilizes deformable 3D convolutions for non-rigid registration to minimize spatial shifts, while the second module enhances contrast and detail. Leveraging the advanced capabilities of the 3D Mamba architecture, experimental results demonstrate that MAMAT outperforms state-of-the-art learning-based methods, achieving up to a 3\% improvement in visual quality and a 15\% boost in object detection. It not only enhances visualization but also significantly improves object detection accuracy, bridging the gap between visual restoration and the effectiveness of surveillance applications.

Original language	English
Title of host publication	2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Number of pages	6
ISBN (Electronic)	9798331514808, 9798331514815
DOIs	https://doi.org/10.48550/arXiv.2503.17700 https://doi.org/10.1109/AVSS65446.2025.11149942
Publication status	Published - 9 Sept 2025
Event	IEEE International Conference on Advanced Visual and Signal-Based Systems - Tainan, Taiwan Duration: 11 Aug 2025 → 13 Aug 2025 https://sites.google.com/view/avss2025-tw

Publication series

Name	IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)
Publisher	IEEE
ISSN (Print)	2643-6205
ISSN (Electronic)	2643-6213

Conference

Conference	IEEE International Conference on Advanced Visual and Signal-Based Systems
Abbreviated title	AVSS 2025
Country/Territory	Taiwan
City	Tainan
Period	11/08/25 → 13/08/25
Internet address	https://sites.google.com/view/avss2025-tw

Keywords

cs.CV

Access to Document

Full-text PDF (accepted manuscript)
This is the accepted author manuscript (AAM) of the article which has been made Open Access under the University of Bristol's Scholarly Works Policy. The final published version (Version of Record) can be found on the publisher's website. The copyright of any third-party content, such as images, remains with the copyright holder.
Accepted author manuscript, 2.17 MBLicence: CC BY

Handle.net

Persistent link

Visual-based Object Recognition under Heat Haze Environment
Anantrasirichai, P. (Principal Investigator)
1/06/23 → 30/05/26
Project: Research, Parent

Cite this

Hill, P., Liu, Z., & Anantrasirichai, N. (2025). Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability. In 2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) Article 11149942 (IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)). Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.48550/arXiv.2503.17700, https://doi.org/10.1109/AVSS65446.2025.11149942

Hill, Paul ; Liu, Zhiming ; Anantrasirichai, Nantheera. / Mamat : 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability. 2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS). Institute of Electrical and Electronics Engineers (IEEE), 2025. (IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)).

@inproceedings{506a69ec0cec4f5dbe116280e2a757db,

title = "Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability",

abstract = "Restoration and enhancement are essential for improving the quality of videos captured under atmospheric turbulence conditions, aiding visualization, object detection, classification, and tracking in surveillance systems. In this paper, we introduce a novel Mamba-based method, the 3D Mamba-Based Atmospheric Turbulence Removal (MAMAT), which employs a dual-module strategy to mitigate these distortions. The first module utilizes deformable 3D convolutions for non-rigid registration to minimize spatial shifts, while the second module enhances contrast and detail. Leveraging the advanced capabilities of the 3D Mamba architecture, experimental results demonstrate that MAMAT outperforms state-of-the-art learning-based methods, achieving up to a 3\textbackslash{}\% improvement in visual quality and a 15\textbackslash{}\% boost in object detection. It not only enhances visualization but also significantly improves object detection accuracy, bridging the gap between visual restoration and the effectiveness of surveillance applications.",

keywords = "cs.CV",

author = "Paul Hill and Zhiming Liu and Nantheera Anantrasirichai",

year = "2025",

month = sep,

day = "9",

doi = "10.48550/arXiv.2503.17700",

language = "English",

series = "IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

booktitle = "2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)",

address = "United States",

note = "IEEE International Conference on Advanced Visual and Signal-Based Systems, AVSS 2025 ; Conference date: 11-08-2025 Through 13-08-2025",

url = "https://sites.google.com/view/avss2025-tw",

}

Hill, P, Liu, Z & Anantrasirichai, N 2025, Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability. in 2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)., 11149942, IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS), Institute of Electrical and Electronics Engineers (IEEE), IEEE International Conference on Advanced Visual and Signal-Based Systems, Tainan, Taiwan, 11/08/25. https://doi.org/10.48550/arXiv.2503.17700, https://doi.org/10.1109/AVSS65446.2025.11149942

Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability. / Hill, Paul; Liu, Zhiming; Anantrasirichai, Nantheera.
2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS). Institute of Electrical and Electronics Engineers (IEEE), 2025. 11149942 (IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)).

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

TY - GEN

T1 - Mamat

T2 - IEEE International Conference on Advanced Visual and Signal-Based Systems

AU - Hill, Paul

AU - Liu, Zhiming

AU - Anantrasirichai, Nantheera

PY - 2025/9/9

Y1 - 2025/9/9

N2 - Restoration and enhancement are essential for improving the quality of videos captured under atmospheric turbulence conditions, aiding visualization, object detection, classification, and tracking in surveillance systems. In this paper, we introduce a novel Mamba-based method, the 3D Mamba-Based Atmospheric Turbulence Removal (MAMAT), which employs a dual-module strategy to mitigate these distortions. The first module utilizes deformable 3D convolutions for non-rigid registration to minimize spatial shifts, while the second module enhances contrast and detail. Leveraging the advanced capabilities of the 3D Mamba architecture, experimental results demonstrate that MAMAT outperforms state-of-the-art learning-based methods, achieving up to a 3\% improvement in visual quality and a 15\% boost in object detection. It not only enhances visualization but also significantly improves object detection accuracy, bridging the gap between visual restoration and the effectiveness of surveillance applications.

AB - Restoration and enhancement are essential for improving the quality of videos captured under atmospheric turbulence conditions, aiding visualization, object detection, classification, and tracking in surveillance systems. In this paper, we introduce a novel Mamba-based method, the 3D Mamba-Based Atmospheric Turbulence Removal (MAMAT), which employs a dual-module strategy to mitigate these distortions. The first module utilizes deformable 3D convolutions for non-rigid registration to minimize spatial shifts, while the second module enhances contrast and detail. Leveraging the advanced capabilities of the 3D Mamba architecture, experimental results demonstrate that MAMAT outperforms state-of-the-art learning-based methods, achieving up to a 3\% improvement in visual quality and a 15\% boost in object detection. It not only enhances visualization but also significantly improves object detection accuracy, bridging the gap between visual restoration and the effectiveness of surveillance applications.

KW - cs.CV

U2 - 10.48550/arXiv.2503.17700

DO - 10.48550/arXiv.2503.17700

M3 - Conference Contribution (Conference Proceeding)

T3 - IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)

BT - 2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)

PB - Institute of Electrical and Electronics Engineers (IEEE)

Y2 - 11 August 2025 through 13 August 2025

ER -

Hill P, Liu Z, Anantrasirichai N. Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability. In 2025 IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS). Institute of Electrical and Electronics Engineers (IEEE). 2025. 11149942. (IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)). doi: 10.48550/arXiv.2503.17700, 10.1109/AVSS65446.2025.11149942

Mamat: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability

Abstract

Publication series

Conference

Keywords

Access to Document

Handle.net

Fingerprint

Projects

Visual-based Object Recognition under Heat Haze Environment

Cite this