View in EDS HTML Full Text PDF Full Text

MAFT: A Lightweight Network for Martian Rock Segmentation Based on an Adaptive Frequency Transformer.

Saved in:
Bibliographic Details
Title: MAFT: A Lightweight Network for Martian Rock Segmentation Based on an Adaptive Frequency Transformer.
Authors: Li, Chu1,2 (AUTHOR), Jia, Yutong1,2 (AUTHOR) jiayutong@hgd.edu.cn, Wan, Gang1,2,3 (AUTHOR), Ma, Qifang1,2,4 (AUTHOR), Liu, Jia1,4 (AUTHOR), Wang, Yang1,2 (AUTHOR), Wang, Biao3 (AUTHOR), Wei, Zhanji1 (AUTHOR)
Source: Remote Sensing. Jun2026, Vol. 18 Issue 11, p1794. 29p.
Subjects: Image segmentation, Transformer models, Mars rovers, Obstacle avoidance (Robotics), Artificial neural networks, Dust storms
Abstract: Highlights: What are the main findings? We proposed the Mars Adaptive Frequency Transformer (MAFT), a lightweight network building upon AFFormer with AKConv and EMCA, which achieves 88.90% Intersection over Union (IoU) with only 2.97 M parameters and 15.49 G floating-point operations (FLOPs), surpassing all compared lightweight and Mars-specific segmentation models. We constructed the TWMARS-V2 dataset with fine-grained annotations, addressing the high omission rate of small rocks in existing datasets and establishing a robust evaluation benchmark. What are the implications of the main findings? With a high inference speed of 35.25 frames per second (FPS) and low computational cost, MAFT is highly suitable for deployment on resource-constrained onboard hardware, enabling real-time obstacle avoidance for future Mars rovers. The network's robustness under dust coverage and complex textures supports automated rock size and morphology statistics through a practical measurement workflow. The segmentation of rocks on the Martian surface is crucial for navigation and obstacle avoidance by Mars rovers. However, frequent dust storms degrade rock surface textures, and the wide range of rock scales—from sub-meter to ten-meter—further complicates segmentation, especially under the strict computational constraints of rover hardware. This paper proposes a lightweight network named MAFT, specifically designed for Martian rock segmentation. The network builds upon the Adaptive Frequency Transformer (AFFormer) and constructs an improved backbone termed the Improved Adaptive Frequency Transformer (IAFFormer). By replacing the traditional self-attention mechanism with a frequency-domain approach, it captures global feature dependencies while reducing the computational complexity from quadratic to linear. The spatially isolated 1 × 1 convolutions in the pixel descriptor module are further replaced with Adaptive Kernel Convolution (AKConv), enabling the backbone to dynamically adjust its sampling positions to conform to the irregular and diverse morphologies of Martian rocks. An Enhanced Multidimensional Convolutional Attention (EMCA) module is introduced as the decoding structure. By integrating max-pooling in the squeeze stage and adaptive dilated convolutions in the excitation stage, EMCA strengthens the boundary perception and long-range dependency modeling of dust-covered rocks without increasing the parameter count. Additionally, we constructed a dataset of Martian rocks for the Zhurong rover (TWMARS-V2) and conducted experiments using a synthetic dataset (SynMars) and a real dataset (MarsData-V2). Experimental results demonstrate that MAFT achieves the highest segmentation accuracy among all compared methods, with only 2.97 M parameters and 15.49 G FLOPs. On the TWMARS-V2 dataset, Pixel Accuracy (PA) reaches 98.17%, and IoU reaches 88.90%. [ABSTRACT FROM AUTHOR]
Copyright of Remote Sensing is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: egs
DbLabel: Engineering Source
An: 194587015
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: MAFT: A Lightweight Network for Martian Rock Segmentation Based on an Adaptive Frequency Transformer.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Li%2C+Chu%22">Li, Chu</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Jia%2C+Yutong%22">Jia, Yutong</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<i> jiayutong@hgd.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Wan%2C+Gang%22">Wan, Gang</searchLink><relatesTo>1,2,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Ma%2C+Qifang%22">Ma, Qifang</searchLink><relatesTo>1,2,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Liu%2C+Jia%22">Liu, Jia</searchLink><relatesTo>1,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Yang%22">Wang, Yang</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Biao%22">Wang, Biao</searchLink><relatesTo>3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wei%2C+Zhanji%22">Wei, Zhanji</searchLink><relatesTo>1</relatesTo> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Remote+Sensing%22">Remote Sensing</searchLink>. Jun2026, Vol. 18 Issue 11, p1794. 29p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Transformer+models%22">Transformer models</searchLink><br /><searchLink fieldCode="DE" term="%22Mars+rovers%22">Mars rovers</searchLink><br /><searchLink fieldCode="DE" term="%22Obstacle+avoidance+%28Robotics%29%22">Obstacle avoidance (Robotics)</searchLink><br /><searchLink fieldCode="DE" term="%22Artificial+neural+networks%22">Artificial neural networks</searchLink><br /><searchLink fieldCode="DE" term="%22Dust+storms%22">Dust storms</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Highlights: What are the main findings? We proposed the Mars Adaptive Frequency Transformer (MAFT), a lightweight network building upon AFFormer with AKConv and EMCA, which achieves 88.90% Intersection over Union (IoU) with only 2.97 M parameters and 15.49 G floating-point operations (FLOPs), surpassing all compared lightweight and Mars-specific segmentation models. We constructed the TWMARS-V2 dataset with fine-grained annotations, addressing the high omission rate of small rocks in existing datasets and establishing a robust evaluation benchmark. What are the implications of the main findings? With a high inference speed of 35.25 frames per second (FPS) and low computational cost, MAFT is highly suitable for deployment on resource-constrained onboard hardware, enabling real-time obstacle avoidance for future Mars rovers. The network's robustness under dust coverage and complex textures supports automated rock size and morphology statistics through a practical measurement workflow. The segmentation of rocks on the Martian surface is crucial for navigation and obstacle avoidance by Mars rovers. However, frequent dust storms degrade rock surface textures, and the wide range of rock scales—from sub-meter to ten-meter—further complicates segmentation, especially under the strict computational constraints of rover hardware. This paper proposes a lightweight network named MAFT, specifically designed for Martian rock segmentation. The network builds upon the Adaptive Frequency Transformer (AFFormer) and constructs an improved backbone termed the Improved Adaptive Frequency Transformer (IAFFormer). By replacing the traditional self-attention mechanism with a frequency-domain approach, it captures global feature dependencies while reducing the computational complexity from quadratic to linear. The spatially isolated 1 × 1 convolutions in the pixel descriptor module are further replaced with Adaptive Kernel Convolution (AKConv), enabling the backbone to dynamically adjust its sampling positions to conform to the irregular and diverse morphologies of Martian rocks. An Enhanced Multidimensional Convolutional Attention (EMCA) module is introduced as the decoding structure. By integrating max-pooling in the squeeze stage and adaptive dilated convolutions in the excitation stage, EMCA strengthens the boundary perception and long-range dependency modeling of dust-covered rocks without increasing the parameter count. Additionally, we constructed a dataset of Martian rocks for the Zhurong rover (TWMARS-V2) and conducted experiments using a synthetic dataset (SynMars) and a real dataset (MarsData-V2). Experimental results demonstrate that MAFT achieves the highest segmentation accuracy among all compared methods, with only 2.97 M parameters and 15.49 G FLOPs. On the TWMARS-V2 dataset, Pixel Accuracy (PA) reaches 98.17%, and IoU reaches 88.90%. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Remote Sensing is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=194587015
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/rs18111794
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 29
        StartPage: 1794
    Subjects:
      – SubjectFull: Image segmentation
        Type: general
      – SubjectFull: Transformer models
        Type: general
      – SubjectFull: Mars rovers
        Type: general
      – SubjectFull: Obstacle avoidance (Robotics)
        Type: general
      – SubjectFull: Artificial neural networks
        Type: general
      – SubjectFull: Dust storms
        Type: general
    Titles:
      – TitleFull: MAFT: A Lightweight Network for Martian Rock Segmentation Based on an Adaptive Frequency Transformer.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Li, Chu
      – PersonEntity:
          Name:
            NameFull: Jia, Yutong
      – PersonEntity:
          Name:
            NameFull: Wan, Gang
      – PersonEntity:
          Name:
            NameFull: Ma, Qifang
      – PersonEntity:
          Name:
            NameFull: Liu, Jia
      – PersonEntity:
          Name:
            NameFull: Wang, Yang
      – PersonEntity:
          Name:
            NameFull: Wang, Biao
      – PersonEntity:
          Name:
            NameFull: Wei, Zhanji
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 06
              Text: Jun2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 20724292
          Numbering:
            – Type: volume
              Value: 18
            – Type: issue
              Value: 11
          Titles:
            – TitleFull: Remote Sensing
              Type: main
ResultId 1