A Comprehensive Review of Fish Feeding Behavior Analysis in Aquaculture: Tasks, Techniques, and Applications

Daoliang Li; Haihua Wang; Jiayin Zhao; Mingyuan Yao; Shulong Zhang; Yingyi Chen

arxiv: 2502.15311 · v3 · submitted 2025-02-21 · 💻 cs.CV · cs.ET

A Comprehensive Review of Fish Feeding Behavior Analysis in Aquaculture: Tasks, Techniques, and Applications

Shulong Zhang , Daoliang Li , Jiayin Zhao , Mingyuan Yao , Yingyi Chen , Haihua Wang This is my paper

Pith reviewed 2026-05-23 02:54 UTC · model grok-4.3

classification 💻 cs.CV cs.ET

keywords fish feeding behavior analysisaquaculturecomputer visionacousticssensorsmultimodal fusionintelligent feedingprecision aquaculture

0 comments

The pith

Fish feeding behavior analysis splits into two subtasks supported by vision, sound, sensors and combined methods.

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

The paper sets out to organize the study of how fish eat in farms into a clear structure. It separates the work into two main parts: identifying feeding actions and measuring how much fish consume. It then reviews different monitoring methods including cameras, underwater sound recording, physical sensors, and ways to combine them, noting where each works best. This organization matters because it can guide better automated feeding systems that use feed more efficiently and reduce waste in fish farming. The review also lists open problems like handling murky water and adapting to different fish species.

Core claim

Fish feeding behavior analysis consists of two core subtasks whose behavioral characteristics and metrics are summarized, with development paths of computer vision, acoustics, sensors, and multimodal fusion technologies analyzed for their advantages, limitations, and scenarios, leading to applications in intelligent feeding and aquaculture management.

What carries the argument

The two core subtasks of fish feeding behavior analysis, which separate the identification of feeding actions from the quantification of feeding intensity.

If this is right

Task standardization allows consistent evaluation across different studies and technologies.
Selection of appropriate sensing methods based on environmental conditions improves system reliability.
Application in closed-loop feeding reduces feed waste and environmental impact.
Addressing listed challenges like multimodal collaboration will advance precision aquaculture.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The review's framework could serve as a basis for developing standardized datasets for fish behavior analysis.
Future work might test the multimodal approaches in real commercial farms to validate the summarized scenarios.
Links between feeding analysis and broader sustainability goals in aquaculture could be explored further.

Load-bearing premise

The existing studies on fish feeding behavior provide a sufficiently complete and unbiased sample of the field's current state.

What would settle it

A new study that identifies major gaps in the reviewed literature or shows that one of the four technologies has been fundamentally mischaracterized in terms of its limitations.

Figures

Figures reproduced from arXiv: 2502.15311 by Daoliang Li, Haihua Wang, Jiayin Zhao, Mingyuan Yao, Shulong Zhang, Yingyi Chen.

**Figure 2.** Figure 2: Flowchart of fish feeding behavior recognition and intensity quantification based on acoustic technology. [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗

**Figure 3.** Figure 3: Structure of the sensor-based fish feeding behavior analysis system. 5.1.1 Based on biosensors Biosensors are surgically inserted into the abdominal cavity or immobilized on the body surface to continuously monitor the fish behaviors and physiological parameters such as heart rate, temperature, orientation and acceleration over time (Brijs et al., 2021), and changes in these parameters during feeding diffe… view at source ↗

**Figure 4.** Figure 4: Structure of multimodal fusion. Multimodal fusion has demonstrated superior performance in detecting and recognizing fish feeding behaviors as it provides richer information compared to single-modality approaches. Qu et al. (2021) explored the correlation between feeding behavior and acoustic signals in Largemouth bass under different culture densities using an acoustic-visual synchronization method. But t… view at source ↗

read the original abstract

Fish feeding behavior analysis is a key foundation for intelligent feeding and precision aquaculture management, and plays an important role in improving feed utilization efficiency, reducing production costs, and mitigating environmental burden. Existing reviews mainly focus on specific technical modalities or related applications in smart aquaculture, which makes it difficult to present the overall development of fish feeding behavior analysis in a comprehensive manner. To address these issues, this paper provides a thematic review of fish feeding behavior analysis in aquaculture, and systematically examines its task definition, technical support, and application status. First, from the task perspective, two core subtasks of fish feeding behavior analysis are clearly distinguished, and relevant behavioral characteristics and evaluation metrics are summarized. Second, from the technical perspective, the development trajectories of computer vision, acoustics, sensors, and multimodal fusion technologies are examined, and their advantages, limitations, and applicable scenarios are analyzed. On this basis, the application value of fish feeding behavior analysis in intelligent feeding and aquaculture management is further summarized. Finally, this paper discusses the challenges in robust perception under complex environments, generalization across fish species and farming scenarios, collaborative multimodal modeling and lightweight deployment, closed loop intelligent feeding, coordinated optimization of multiple tasks, and long-term production validation, and outlines future research directions. This review provides a reference for task standardization, technical selection, and engineering application in fish feeding behavior analysis, and offers insights into the development of smart aquaculture and sustainable aquaculture management.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This is a standard thematic review that maps tasks and tech in fish feeding analysis but rests on an unstated literature sample and adds no new empirical or methodological ground.

read the letter

The core takeaway is that the paper pulls together existing work on fish feeding behavior into two subtasks, then walks through computer vision, acoustics, sensors, and fusion approaches with their trade-offs and use cases in aquaculture. It also flags practical challenges like robustness in messy environments and cross-species generalization. That structure is the main thing it offers: a single document that someone entering the area can use to see the landscape without hunting down dozens of papers themselves. It does this cleanly enough on the surface, with clear sections on metrics, advantages, and application value in intelligent feeding systems. No new data or derivations appear, which is expected for a review, but the organization itself is straightforward and could save time for practitioners choosing sensors or models. The soft spot is the missing search methodology. The abstract and stress-test note both flag the absence of databases, keywords, date ranges, or inclusion rules, so it is impossible to judge whether the cited work is representative or just the authors' familiar slice. If the full text does not add that detail, the claims about field-wide trajectories and challenges rest on an opaque sample. Minor issues include the usual review risk of uneven depth across modalities, but nothing load-bearing shows up from what is visible. This paper is for aquaculture engineers and precision-farming researchers who need a quick reference rather than for theorists or method developers. A reader already deep in the subfield will probably not find surprises. It deserves a serious referee if the full manuscript supplies the search protocol and a balanced citation list; without those, it is closer to a useful internal note than a citable synthesis. I would send it for review with a request to document the literature selection process explicitly.

Referee Report

1 major / 2 minor

Summary. The manuscript provides a thematic review of fish feeding behavior analysis in aquaculture. It distinguishes two core subtasks from the task perspective, examines development trajectories and trade-offs of computer vision, acoustics, sensors, and multimodal fusion from the technical perspective, summarizes application value in intelligent feeding and aquaculture management, and outlines challenges (robust perception, cross-species generalization, multimodal modeling, lightweight deployment, closed-loop feeding, multi-task optimization, long-term validation) together with future directions.

Significance. If the reviewed literature is representative, the structured organization of subtasks, metrics, technology trajectories, and application scenarios would supply a useful reference for task standardization and technical selection in precision aquaculture.

major comments (1)

[Introduction / Thematic Review Approach] The central claim that the paper delivers a 'comprehensive' thematic review is load-bearing on the representativeness of the selected literature, yet no literature-search methodology (databases, keywords, date range, inclusion/exclusion criteria, or PRISMA-style flow) is stated in the introduction or methods. This omission prevents verification that the distinctions between the two subtasks and the summarized advantages/limitations of each modality reflect the actual state of the field rather than an uncharacterized sample.

minor comments (2)

[Task Perspective] Clarify the exact definitions and evaluation metrics of the two core subtasks in a dedicated table or subsection to improve traceability for readers.
[Technical Perspective] Ensure that every cited work appears in the reference list and that the reference list is complete; several modality-specific claims would benefit from explicit citation anchors.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on our thematic review. The single major comment is addressed point-by-point below. We agree that greater transparency on literature selection will strengthen the manuscript and will revise accordingly.

read point-by-point responses

Referee: [Introduction / Thematic Review Approach] The central claim that the paper delivers a 'comprehensive' thematic review is load-bearing on the representativeness of the selected literature, yet no literature-search methodology (databases, keywords, date range, inclusion/exclusion criteria, or PRISMA-style flow) is stated in the introduction or methods. This omission prevents verification that the distinctions between the two subtasks and the summarized advantages/limitations of each modality reflect the actual state of the field rather than an uncharacterized sample.

Authors: We acknowledge the validity of this observation. The manuscript presents a thematic review organized around task definitions, modality trajectories, and applications rather than a systematic review; literature was selected for its direct relevance to the two core subtasks, the four technical categories, and the listed challenges. No formal search protocol or PRISMA diagram was included. To improve verifiability while preserving the thematic focus, we will insert a concise subsection (or paragraph) in the Introduction that states the primary databases consulted (IEEE Xplore, Web of Science, Google Scholar), representative keywords, approximate date range, and inclusion rationale. This addition will allow readers to assess coverage without converting the paper into a systematic review. We believe the distinctions drawn between subtasks and the modality trade-offs remain grounded in the cited works and will not require substantive alteration beyond this transparency measure. revision: yes

Circularity Check

0 steps flagged

No circularity: pure literature synthesis with no derivation chain

full rationale

This is a thematic review paper that summarizes tasks, techniques, and applications from existing literature on fish feeding behavior analysis. It contains no equations, fitted parameters, predictions, uniqueness theorems, or mathematical derivations of any kind. All content consists of narrative synthesis of prior work, with no self-referential reductions where a claimed result is equivalent to its inputs by construction. The absence of any load-bearing derivation chain means none of the enumerated circularity patterns apply.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is a review paper with no new mathematical derivations, empirical fits, or postulated entities. No free parameters, axioms, or invented entities are introduced.

pith-pipeline@v0.9.0 · 5802 in / 1161 out tokens · 23444 ms · 2026-05-23T02:54:49.069054+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

F., & Darozat, F

https://doi.org/10.1038/s42256-023-00633-5 Subakti, A., Khotimah, Z. F., & Darozat, F. M. (2017). Preliminary study of acceleration based sensor to record nile tilapia ( Oreochromis niloticus ) feeding behavior at water surface. Journal of Physics: Conference Series, 795(1), 012060. https://doi.org/10.1088/1742-6596/795/1/012060 Sun, L., Wang, X., Wang, B...

work page doi:10.1038/s42256-023-00633-5 2017
[2]

https://doi.org/10.16535/j.cnki.dlhyxb.2022-307 Xu L., Huang X., & Liu S. (2022). Recognition of fish feeding intensity based on improved LRCN. Transactions of the Chinese Society for Agricultural Machinery, 53(10), 236–241. Xu, L., Huang, Z., Long, W., Jiang, L., & Tong, X. (2024). Classification Model of Fish Feeding Intensity Based on MobileViT CBAM Bi...

work page doi:10.16535/j.cnki.dlhyxb.2022-307 2022

[1] [1]

F., & Darozat, F

https://doi.org/10.1038/s42256-023-00633-5 Subakti, A., Khotimah, Z. F., & Darozat, F. M. (2017). Preliminary study of acceleration based sensor to record nile tilapia ( Oreochromis niloticus ) feeding behavior at water surface. Journal of Physics: Conference Series, 795(1), 012060. https://doi.org/10.1088/1742-6596/795/1/012060 Sun, L., Wang, X., Wang, B...

work page doi:10.1038/s42256-023-00633-5 2017

[2] [2]

https://doi.org/10.16535/j.cnki.dlhyxb.2022-307 Xu L., Huang X., & Liu S. (2022). Recognition of fish feeding intensity based on improved LRCN. Transactions of the Chinese Society for Agricultural Machinery, 53(10), 236–241. Xu, L., Huang, Z., Long, W., Jiang, L., & Tong, X. (2024). Classification Model of Fish Feeding Intensity Based on MobileViT CBAM Bi...

work page doi:10.16535/j.cnki.dlhyxb.2022-307 2022