Pith. sign in

REVIEW 4 major objections 5 minor 19 references

Lightweight Shrimp Disease Detection Research Based on YOLOv8n

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A YOLOv8n variant with a reparameterized detection head, a multi-scale EMCM module, and an improved attention mechanism reports 92.7% mAP@0.5 on a three-class shrimp disease dataset while cutting parameters by 32.3%.

desk verdict A routine YOLOv8n tweak with a genuinely new shrimp-disease dataset, but the reported gains sit on a thin single-run evaluation and one implausible ablation row. read the letter →

arxiv 2507.02354 v1 pith:254LK5IQ submitted 2025-07-03 cs.CV

Fei Yuhuan , Wang Gengchen , Liu Fenghao , Zang Ran , Sun Xufei , Chang Hao This is my paper · ORCID
classification cs.CV
keywords shrimpdiseasedetectionYOLOv8nlightweightobjectreparameterizedheadmulti-scaleconvolutionattentionmechanismURPC2020mAP@0.5
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that a standard YOLOv8n detector can be made simultaneously lighter and more accurate for shrimp disease detection. The proposed network swaps in a reparameterized lightweight detection head (RLDD), replaces standard bottleneck convolutions with a multi-scale EMCM module, and adds an improved SegNext-style attention mechanism. On a self-collected dataset of 1,047 images covering white spot, black gill, and yellow gill disease, the model reports 92.7% mAP@0.5 with 2.1 million parameters — a 3 percentage point gain over YOLOv8n at a 32.3% parameter reduction. If the result holds, it means accurate disease screening could run on low-cost, resource-constrained devices in shrimp farms rather than requiring cloud servers.

What carries the argument

The load-bearing mechanism is the combination of three architectural substitutions. RLDD uses reparameterized convolutions: during training the detection head has 3×3, 1×1, and 3×3 average-pool branches; at inference these merge losslessly into one 3×3 convolution, cutting parameters while keeping multi-scale capacity. EMCM splits input channels, passes one branch straight through as original features and the other through parallel 3×3 and 5×5 convolutions, then concatenates and compresses with a 1×1 convolution; replacing the standard Bottleneck in selected C2f blocks gives multi-scale receptive fields. The SegNext_Attention module builds an attention map from depthwise and multi-scale depthwise convolutions and multiplies it onto the feature map, letting the model focus on disease-relevant regions.

What would settle it

Re-train the baseline and the full model on the same 1,047 images under several random 9:1 splits (or on an independent public shrimp-disease dataset) and compare the mAP differences; if the 92.7% versus 89.7% gap shrinks below the run-to-run standard deviation, the claimed improvement is not established.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that a YOLOv8n detector can be made lighter and more accurate for shrimp disease detection by stacking three changes: an RLDD head that trains as a multi-branch structure and collapses to a single 3×3 convolution at inference; a C2f-EMCM module that splits channels and extracts 3×3 and 5×5 multi-scale features; and an improved SegNext_Attention mechanism that computes pixel-level attention to suppress background. The full system reports 92.7% mAP@0.5, 87.5% precision, and 88.9% recall on the self-built dataset with 2.1M parameters, versus 89.7% mAP and 3.1M parameters for YOLOv8n; on URPC2020 it reports 87.2% mAP against 83.1% for YOLOv8n. The paper concludes that the architecture achieves the desired accuracy-efficiency balance for shrimp aquaculture monitoring.

Load-bearing premise

The reported improvements depend on the single 9:1 split of a private 1,047-image dataset, with no held-out test set and no repeated runs, so the 3-point mAP gain could change if the data or split changes.

Editorial extensions

If this is right

  • The model can be deployed on edge devices for near-real-time shrimp disease screening: 2.1M parameters and a 32.3% parameter cut relative to YOLOv8n at equal or better mAP.
  • The RLDD reparameterized head alone yields a 25.8% parameter reduction with a precision gain, so reparameterization of the detection head is a generally applicable lightweighting technique.
  • The architecture transfers to a different underwater detection domain (URPC2020) with a 4.1-point mAP gain over YOLOv8n, suggesting the module combination generalizes beyond shrimp disease.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's evidence would be stronger with repeated runs; absent that, the practical takeaway should be treated as an existence proof that the three-module stack can work, not as a guaranteed improvement on a new farm's dataset.
  • Because the EMCM+SegNext ablation lands exactly at baseline mAP while the full stack improves, the modules appear to interact rather than simply add; testing each on more varied data could clarify where the synergy comes from.
  • The URPC2020 result would be more compelling if the protocol (zero-shot versus fine-tuned) were explicit, since zero-shot generalization across domains would be the stronger claim.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes three modifications to YOLOv8n for shrimp disease detection: a reparameterized lightweight detection head (RLDD), an efficient multi-scale convolution module (C2f-EMCM), and an improved SegNext_Attention self-attention module. On a self-constructed dataset of 1047 images containing three shrimp disease classes, the authors report mAP@0.5 of 92.7% with 2.1M parameters, compared with 89.7% and 3.1M for the YOLOv8n baseline, and 87.2% mAP on the URPC2020 dataset. The paper's evidence is entirely experimental: an eight-row ablation study, comparisons with several YOLO variants and lightweight backbones, qualitative detection visualizations, and a two-model generalization experiment on URPC2020.

Significance. If the reported numbers are reproducible, the combination of a 3.0-point mAP gain with a 32.3% parameter reduction is practically useful for edge deployment in aquaculture monitoring, and the improved URPC2020 result would be an encouraging sign of cross-domain robustness. The paper also has a reasonable structure: each proposed component is described, an ablation is attempted, and the final model is compared with six alternatives. However, the empirical support is currently too thin to certify the central claim. The results rest on a single run over one small private train/validation split, the ablation contains an unexplained exact duplication of the baseline, and the URPC2020 protocol is not specified. These gaps make the headline accuracy and efficiency claims unverifiable from the manuscript as written.

major comments (4)
  1. [§3.1.2 / §3.2.1] The entire accuracy claim rests on a single training run with a 9:1 random split of 1047 images, and the manuscript reports no separate test set, no variance across random seeds, and no confidence intervals. With roughly 105 validation images, the reported 3.0-point mAP gain and 9.1-point precision gain over YOLOv8n could be within sampling noise. Please report mean and standard deviation over at least three seeds, and preferably use a fixed held-out test split that is never used for model selection.
  2. [§3.2.1, Table 3] Experiment 7 (EMCM + SegNext_Attention) reports Precision 78.4% and mAP@0.5 89.7%, exactly equal to the baseline experiment 1, even though experiments 3 and 4 individually improve both metrics. If this exact duplication is not a copying or recording error, it needs a mechanistic explanation; as presented, it undermines the compositional interpretation of the final model and casts doubt on the reliability of the ablation table.
  3. [§3.2.4] The generalization experiment on URPC2020 does not state whether the model was fine-tuned on URPC2020 or evaluated zero-shot, how many training epochs and what hyperparameters were used, or how the detection head was adapted to the different class set (starfish, sea urchin, sea cucumber, etc.). Without this information, the reported 87.2% mAP and the 4.1-point improvement over YOLOv8n are not reproducible and cannot be interpreted as evidence of domain generalization.
  4. [§3.2.2 / Abstract] The paper repeatedly claims reduced 'computational complexity' and superiority in 'model size,' but Table 4 reports only parameter counts and does not include FLOPs, inference time, or model file size. Please add these quantities for all compared models; otherwise the efficiency part of the central claim is only partially supported.
minor comments (5)
  1. [§3.1.3] Equation (6) appears to define AP per class but refers to 'n' as the number of classes; please clarify the interpolation formula and use consistent notation for the class index and the number of recall intervals.
  2. [§3.1.1] The experimental environment lists 'PyTorch 1.19.0,' which is not a released version of PyTorch; please correct this or specify the exact version used.
  3. [General] The text jumps from Table 1 to Table 3; there is no Table 2. Please renumber the tables or add the missing table.
  4. [§2.3] Equations (1) and (2) are poorly typeset: the summation argument, the convolution notation, and the element-wise multiplication symbol are unclear. Please rewrite these equations with standard mathematical notation.
  5. [§3.2.4] The visualization in Figure 7 is described as showing improvements on specific categories, but no per-class AP values are reported for URPC2020; please include per-class results to substantiate the qualitative claims.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported gains are empirical measurements against external baselines and an external benchmark, not constructions derived from the paper's own definitions.

full rationale

The paper's central claims are experimental: a 92.7% mAP@0.5 on a self-constructed shrimp disease dataset, a 32.3% parameter reduction relative to YOLOv8n, and a 4.1% mAP gain on URPC2020. These are presented as measured results of training and evaluation, not as consequences of definitions or fitted parameters. No equation in the paper reduces a predicted quantity to an input by construction, and no load-bearing step depends on a self-citation; references to prior work are external baselines or standard components rather than a chain of the authors' own unpublished claims. The ablation table does contain an unexplained exact reproduction of the baseline mAP in experiment 7, and the evaluation uses a single 9:1 split of 1047 images with no variance reporting, but those are methodological and reproducibility concerns, not circularity. The model's gains are not forced by the paper's own formalism, so the appropriate circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 3 invented entities

No explicit free parameters beyond standard training hyperparameters are introduced. The model's learned weights are outputs, not input parameters. The central claim relies on the correctness of the private dataset labels, the validity of YOLOv8n as a baseline, the SegNext_Attention implementation from the cited literature, and the assumption that URPC2020 is an appropriate external benchmark.

assumptions (4)
  • domain assumption The YOLOv8n baseline is valid and its measured performance (89.7% mAP@0.5) is reproducible under the paper's settings.
    Section 3.2.1 Table 3 treats YOLOv8n as the reference, and all comparisons are relative to it.
  • domain assumption The SegNext_Attention module and its components (MSCA, Hamburger from [13]) behave as described in the cited literature.
    Section 2.3 adopts the module for feature extraction without re-derivation.
  • domain assumption The self-constructed shrimp disease dataset (1047 images, three classes) has accurate bounding box and class annotations.
    Section 3.1.2 relies on these annotations as ground truth for all metrics.
  • domain assumption URPC2020 is a valid external benchmark for evaluating generalization to underwater detection.
    Section 3.2.4 uses URPC2020 as the generalization test without justifying its relevance to shrimp disease.
invented entities (3)
  • RLDD detection head independent evidence
    purpose: Reduces parameter count in YOLOv8n's decoupled head via reparameterized convolutions while preserving accuracy.
    Ablation experiment 2 shows 2.3M parameters with 91.5% mAP, and URPC2020 results support its use.
  • C2f-EMCM module independent evidence
    purpose: Multi-scale feature extraction with grouped convolutions to replace standard C2f Bottleneck convolutions.
    Ablation experiment 3 shows precision improvement from 78.4% to 86.5% and mAP from 89.7% to 92.1%.
  • Improved SegNext_Attention independent evidence
    purpose: Self-attention to suppress background and focus on disease regions.
    Ablation experiment 4 shows a small mAP gain alone (90.5%), and experiment 6 with RLDD reaches 92.1%, indicating a positive interaction.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Lightweight Shrimp Disease Detection Research Based on YOLOv8n." pith.science (2026). https://pith.science/paper/254LK5IQ

@misc{pith2026250702354,
  author       = {Pith},
  title        = {Pith review of: Lightweight Shrimp Disease Detection Research Based on YOLOv8n},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/254LK5IQ}},
  note         = {Machine review of arXiv:2507.02354}
}
read the original abstract

Shrimp diseases are one of the primary causes of economic losses in shrimp aquaculture. To prevent disease transmission and enhance intelligent detection efficiency in shrimp farming, this paper proposes a lightweight network architecture based on YOLOv8n. First, by designing the RLDD detection head and C2f-EMCM module, the model reduces computational complexity while maintaining detection accuracy, improving computational efficiency. Subsequently, an improved SegNext_Attention self-attention mechanism is introduced to further enhance the model's feature extraction capability, enabling more precise identification of disease characteristics. Extensive experiments, including ablation studies and comparative evaluations, are conducted on a self-constructed shrimp disease dataset, with generalization tests extended to the URPC2020 dataset. Results demonstrate that the proposed model achieves a 32.3% reduction in parameters compared to the original YOLOv8n, with a mAP@0.5 of 92.7% (3% improvement over YOLOv8n). Additionally, the model outperforms other lightweight YOLO-series models in mAP@0.5, parameter count, and model size. Generalization experiments on the URPC2020 dataset further validate the model's robustness, showing a 4.1% increase in mAP@0.5 compared to YOLOv8n. The proposed method achieves an optimal balance between accuracy and efficiency, providing reliable technical support for intelligent disease detection in shrimp aquaculture.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 17 canonical work pages

  1. [1]

    Shri mp bacterial and parasitic disease listed in the OI E: A review[J]

    YU Y B, CHOI J H, KANG J C, et al. Shri mp bacterial and parasitic disease listed in the OI E: A review[J]. Microbial Pathogenesis, 2022, 166: 105545

    work page 2022

  2. [2]

    Viral shr imp diseases listed by the OIE: A review[J]

    LEE D, YU Y B, CHOI J H , et al. Viral shr imp diseases listed by the OIE: A review[J]. Viru ses, 2022, 14(3): 585

    work page 2022

  3. [3]

    You only look once: Unified, real -time object detection[C]//Proceedings of 2016 IEEE Conferenc e on Computer Vision and Pattern Recognition

    REDMON J, DIVV ALA S, GIRSHICK R , et al. You only look once: Unified, real -time object detection[C]//Proceedings of 2016 IEEE Conferenc e on Computer Vision and Pattern Recognition. 2 016: 779 -788. DOI: 10.1109/CVPR.2016.91

    doi:10.1109/cvpr.2016.91 2016

  4. [4]

    改进 YOLOv7 的轻 量化水下目标检测算法 [J]

    辛世澳, 葛海波, 袁昊, 等. 改进 YOLOv7 的轻 量化水下目标检测算法 [J]. 计算机工程与应用, 20 23, 60(3): 88 -99

    work page

  5. [5]

    基于 SimAM 注意力机 制的 DCN-YOLOv5 水下目标检测[J]

    刘向举, 刘洋, 蒋社想. 基于 SimAM 注意力机 制的 DCN-YOLOv5 水下目标检测[J]. 重庆工商大 学学报(自然科学版), 2023, 40(2): 76 -81

    work page 2023

  6. [6]

    YOLOv5s -CA: A modified YOLOv5s network with coordinate att ention for underwater target detection[J]

    WEN G, LI S, LIU F , et al. YOLOv5s -CA: A modified YOLOv5s network with coordinate att ention for underwater target detection[J]. Sensors, 2023, 23(7): 3367

    work page 2023

  7. [7]

    In situ sea cucu mber detection based on deep learning approach[C] //2018 OCEANS MTS/IEEE Kobe Techno -Oceans (OTO)

    XIA C, FU L, LIU H , et al . In situ sea cucu mber detection based on deep learning approach[C] //2018 OCEANS MTS/IEEE Kobe Techno -Oceans (OTO). Kobe, Japan: IEEE, 2018: 1 -4

    work page 2018

  8. [8]

    基于改进 YOLO 卷积神经网络的水下海参检测 [J]

    翟先一, 魏鸿磊, 韩美奇, 等. 基于改进 YOLO 卷积神经网络的水下海参检测 [J]. 江苏农业学报, 2023, 39(7): 1543 -1553

    work page 2023

Show all 19 references
  1. [9]

    An improv ed YOLOv5s -based scheme for target detection in a complex underwater environment[J]

    HOU C, GUAN Z, GUO Z , et al. An improv ed YOLOv5s -based scheme for target detection in a complex underwater environment[J]. Journal of Marine Science and Engineering, 2023, 11(5): 104 1

    2023

  2. [10]

    An improved YOLOV 5 based on triplet attention and prediction head o ptimization for marine organism detection on unde rwater mobile platforms[J]

    LI Y , BAI X, XIA C . An improved YOLOV 5 based on triplet attention and prediction head o ptimization for marine organism detection on unde rwater mobile platforms[J]. Journal of Marine Scie nce and Engineering, 2022, 10(9): 1230

    2022

  3. [11]

    U -YOLOv7: A network for underwater organism detection[J]

    YU G, CAI R, SU J , et al. U -YOLOv7: A network for underwater organism detection[J]. Eco logical Informatics, 2023, 75: 102108

    2023

  4. [12]

    I mproved Underwater Object Detection Algorithm of YOLOv7[J]

    LIANG X M, RAN L I, HAI Y U , et al. I mproved Underwater Object Detection Algorithm of YOLOv7[J]. Journal of Computer Engineering & Applications, 2024, 60(6)

    2024

  5. [13]

    Is at tention better than matrix decomposition?[J]

    GENG Z, GUO M H, CHEN H , et al. Is at tention better than matrix decomposition?[J]. arXiv preprint arXiv:2109.04553, 2021

    2021 arXiv

  6. [14]

    Apple ste m/calyx real -time recognition using YOLO -v5 algo rithm for fruit automatic loading syst em[J]

    WANG Z, JIN L, WANG S , et al. Apple ste m/calyx real -time recognition using YOLO -v5 algo rithm for fruit automatic loading syst em[J]. Postha rvest Biology and Technology, 2022, 185: 111808

    2022

  7. [15]

    Re search on traffic sign detection based on improve d YOLOv8[J]

    HUANG Z, LI L, KRIZECK G C , et al. Re search on traffic sign detection based on improve d YOLOv8[J]. Journal of Computer and Communi cations, 2023, 11(7): 226 -232

    2023

  8. [16]

    YOLOv10: Real -time end -to-end object dete ction[J/OL]

    WANG C Y , BOCHKOVSKIY A, LIAO H Y M . YOLOv10: Real -time end -to-end object dete ction[J/OL]. arXiv preprint arXiv:2305.10530, 202 3

    arXiv

  9. [17]

    YOLOv11: Dynami c feature fusion for real -time detection[J]

    LI X, WANG W, HU X . YOLOv11: Dynami c feature fusion for real -time detection[J]. IEEE T ransactions on Pattern Analysis and Machine Intell igence, 2024, 46(3): 1025 -1038

    2024

  10. [18]

    CHEN J, KAO S H, HE H , et al. Run, don’ t walk: Chasing higher FLOPS for faster neural n 费玉环，等：基于 YOLOv8n 的轻量化虾病害检测研究 9 etworks[C]//Proceedings of 2023 IEEE/CVF Confe rence on Computer Vision and Pattern Recognitio n (CVPR). Vancouver, Canada, 2023: 12021 -1203 1

    2023

  11. [19]

    EfficientDet: Sc alable and efficient object detection[C]//Proceeding s of 2020 IEEE/CVF Conference on Computer Vi sion and Pattern Recognition (CVPR)

    TAN M, PANG R, LE Q V . EfficientDet: Sc alable and efficient object detection[C]//Proceeding s of 2020 IEEE/CVF Conference on Computer Vi sion and Pattern Recognition (CVPR). 2020: 1078 1-10790. Research on Lightweight Shrimp Disease Detection Based on YOLOv8n FEI Yuhuan, WAN...

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.