Pith. sign in

REVIEW 5 major objections 6 minor 46 references

FUME: Fused Unified Multi-Gas Emission Network for Livestock Rumen Acidosis Detection

T0 review · 5 major / 6 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read FUME shows that paired CO2 and CH4 infrared images can classify rumen acidosis states with high accuracy and low computational cost.

desk verdict FUME introduces a genuinely new dual-gas OGI dataset for in vitro rumen acidosis detection and reports strong segmentation results, but the classification claim rests on a dataset design that may allow non-gas shortcuts, and the baseline comparisons look mistuned. read the letter →

arxiv 2601.08205 v1 pith:54WU5G4A submitted 2026-01-13 cs.CV cs.LG

Taminul Islam , Toqi Tahamid Sarker , Mohamed Embaby , Khaled R Ahmed , Amer AbuGhazaleh This is my paper · ORCID
classification cs.CVcs.LG
keywords rumenacidosisopticalgasimagingdual-gasfusionmulti-tasklearningplumesegmentationlivestockhealthmonitoringCO2andCH4emissionsdeep
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

The paper is trying to establish that the gas a cow's rumen fermentation emits, captured as paired CO2 and CH4 infrared images, carries enough information to diagnose acidosis without inserting a probe. It introduces the first dual-gas optical imaging dataset for rumen health and a lightweight multi-task network that both segments the gas plumes pixel-by-pixel and classifies the sample as Healthy, Transitional, or Acidotic. If correct, this would make continuous, non-invasive rumen health monitoring feasible at scale, replacing invasive pH measurement. The reported numbers — 80.99% mIoU, 98.82% accuracy, 1.28M parameters — sit well above prior gas-imaging baselines at roughly a tenth of the compute.

What carries the argument

A lightweight dual-stream encoder (Fast-SCNN backbone with shared weights) extracts CO2 and CH4 features; modality-specific self-attention models long-range plume shape; channel attention fuses the streams by learning which feature channels from each gas should combine; and separate decoders produce segmentation masks while a classifier head outputs the health label. The authors justify channel attention over spatial cross-attention by their finding that the two gases do not align pixel-wise but do share correlated aggregate statistics.

What would settle it

Take the trained model and evaluate it on frames from a new fermentation session held at an intermediate, untrained pH (e.g., 5.75) where the ground-truth class is ambiguous; if classification accuracy collapses toward chance while the model still performs on the original six pH levels, the model likely exploited pH-level-specific visual shortcuts rather than learning a general pH–gas mapping. Alternatively, swap the CO2 and CH4 camera inputs but keep the labels and show that accuracy remains high, which would indicate the model ignored the physiological gas identity.

Watch

Extended reading notes

Core claim

The paper claims to be the first to demonstrate deep-learning-based rumen acidosis detection from dual-gas optical imaging. Its central discovery is that a weight-shared dual-stream network, fusing CO2 and CH4 features through channel attention and trained jointly for gas plume segmentation and health classification, can predict pH-derived health state from in vitro fermentation images. Ablations show CO2 is the primary discriminative gas, CH4 refines boundaries, and the two tasks reinforce each other; attention that is modality-specific plus channel-wise fusion outperforms expensive cross-modal spatial attention.

Load-bearing premise

The assumption that the visual appearance of the gas plumes in the images is causally driven by pH-dependent emission rates (CO2 rising, CH4 falling under acidosis) and not by other differences between the six fermentation conditions, such as tube placement, plume brightness, or camera settings.

Editorial extensions

If this is right

  • If the central claim holds, rumen pH health could be monitored continuously and non-invasively from infrared gas imagery, avoiding rumenocentesis and bolus sensors.
  • The reported efficiency (1.28M parameters, 1.97G MACs, 326 FPS) suggests a single camera unit could run real-time diagnostics on edge hardware.
  • CO2 emerges as the dominant signal, so even a single-gas CO2 camera might provide a practical early-detection device, with CH4 adding boundary precision.
  • The introduced dataset of 8,967 annotated dual-gas frames across six pH levels gives the research community a resource to build on for gas-based health assessment.
  • Dual-task learning — combining segmentation with classification — appears necessary to reach the reported accuracy, pointing to a design principle for future gas-imaging diagnostics.

Reading between the lines

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

  • The design insight that modalities without pixel correspondence should be fused through channel statistics rather than spatial attention could generalize to other sensor pairs (e.g., multispectral, thermal-visible) with correlated but misaligned signals.
  • Because the six pH levels are discrete conditions, the model may have learned a session- or pH-level shortcut rather than a continuous gas–pH relationship; testing on intermediate pH values (e.g., 5.75) or on new fermentation sessions would reveal whether the classifier truly tracks pH.
  • A natural next step the authors leave implicit is temporal modeling: gas plumes are intermittent, so a video-level model that integrates plume appearance over time could improve robustness in live animals.
  • The strong performance gap over single-gas baselines suggests that the combination of CO2 and CH4, even with unequal visibility, encodes more physiological information than either gas alone — a claim testable by measuring gas concentrations directly instead of relying on image appearance.
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

5 major / 6 minor

Summary. The paper introduces FUME, a lightweight dual-stream network that jointly performs gas plume segmentation and rumen health classification from paired CO2 and CH4 optical gas imaging (OGI) frames captured in an in vitro fermentation setup. The method combines a weight-shared Fast-SCNN encoder, modality-specific self-attention, and channel attention fusion. The authors contribute a new dual-gas OGI dataset of 8,967 annotated frames across six pH levels, and report that FUME achieves 80.99% mIoU, 98.82% classification accuracy, 1.28M parameters, and 1.97G MACs, outperforming existing segmentation baselines and OGI-specific methods. Ablations indicate CO2 is the dominant modality and that both tasks are needed for best performance.

Significance. If the claims hold, this is a useful proof-of-concept for non-invasive, gas-imaging-based metabolic monitoring in livestock, with a novel dual-gas dataset and an efficient architecture that could enable real-time edge deployment. The explicit release of a new dataset and the efficient design are strengths. However, the central feasibility claim depends on the model learning pH-related gas emission patterns rather than session- or condition-specific visual shortcuts, and this is not yet convincingly established.

major comments (5)
  1. [§5.1, Table 1] The performance comparison reports single-run metrics without error bars or significance tests. For a claim of 'outperforming state-of-the-art,' this is insufficient, especially because several baselines (CMX 33.65%, RTFNet 46.47%, MTINet 46.37%, DANet 48.50%) are at or near chance (33.3% for three classes). This pattern strongly suggests these baselines were not adequately trained or adapted (e.g., only 20 epochs, no per-model hyperparameter tuning). Please report mean±std over at least 3–5 seeds and provide evidence that baselines were tuned to convergence, or temper the superiority claim accordingly.
  2. [§4.1, Dataset] The dataset consists of only six discrete pH levels, and while splits keep fermentation sessions intact, there is no information about the number of fermentation sessions, how many sessions per pH, or whether a session-wise evaluation was performed. With fixed camera distance, tube geometry, and lighting, any session- or condition-specific visual cue (tube position, plume brightness, background clutter) could provide a shortcut for the 98.82% classification accuracy without the model learning gas-based physiology. The paper's Limitations section (§5.4) discusses in vivo translation but does not address this in vitro confound. Please provide session-level statistics, a session-wise cross-validation, and/or analyses that isolate the contribution of gas plumes versus non-gas image content (e.g., segmentation-masked classification, or controlling for tube/background features).
  3. [§5.3, Table 4] The ablation terminology is ambiguous relative to the final architecture. The proposed FUME already includes modality-specific self-attention and channel attention fusion, yet the table lists '+ Self-Attention Only' as a variant that degrades mIoU by −0.63 pp. It is unclear whether this variant removes channel attention, removes the existing self-attention, or adds an extra attention module. Similarly, 'Segmentation Only' reports 33.65% classification accuracy, but if the classification head is removed, how is this accuracy obtained? Clarify what components are ablated and how accuracy is computed for variants without a classification head.
  4. [§3.3, Eq. (4)] The self-attention module operates on flattened spatial features without any positional encoding. For gas plume segmentation, spatial location and plume morphology are potentially informative, and a purely content-based attention mechanism may be insensitive to spatial layout. Please justify the omission of positional encoding or demonstrate empirically that it does not affect performance.
  5. [Abstract vs. Conclusion] The abstract states 'Codes are available at https://github.com/taminulislam/fume', while the conclusion says 'The dataset and code will be released upon acceptance.' Please reconcile these statements and specify exactly what is currently available / will be released.
minor comments (6)
  1. [§4.2] Training details list 20 epochs and a fixed learning rate for all methods. This may be too few for transformer-based baselines like CMX; consider reporting whether the number of epochs was tuned per model.
  2. [Table 1] The latency measurement uses batch size 1 and FP32; please state whether the reported FPS is the reciprocal of the CUDA-synchronized latency, as the numbers appear inconsistent (e.g., 1000/3.06 ≈ 326.8, which is consistent, but this should be explicit).
  3. [§4.1] The statement 'frames from the same fermentation session were kept within a single split' is important, but the number of sessions is never given. Please add a session-level breakdown to Table 2.
  4. [§5.3] In the ablation results, 'CO2 Only' reduces mIoU by only 0.47 pp relative to FUME, while CH4-only 'fails entirely' for classification. This asymmetry is presented as evidence for complementarity, but the magnitude of the CH4 contribution to segmentation is small; the discussion should be more cautious about the practical value of the CH4 stream.
  5. [§3.4] The multi-task loss weight λ=0.5 is fixed; no sensitivity analysis is provided. Since the classification-segmentation trade-off is emphasized in §5.3, a sweep over λ would strengthen the design justification.
  6. [References] Reference [17] is a systematic review of multi-organ segmentation and may not be the original source for HD95/ASD metrics; please cite the original metric definitions (e.g., Huttenlocher et al. for Hausdorff distance).

Circularity Check

0 steps flagged · score 0.0 of 10

No circular dependency: FUME's predictions are evaluated against external pH labels and manually annotated masks; self-citations are baseline/protocol references, not load-bearing.

full rationale

The paper's derivation chain is empirical rather than definitional. The health labels are assigned from externally measured pH (Section 4.1: 'Samples were maintained at six controlled pH levels...'), and the segmentation targets are manually annotated pixel masks ('Each frame was manually annotated with pixel-level masks'). FUME's classification and segmentation outputs are compared against these external labels and masks, so no fitted quantity is renamed as a prediction. The architecture is assembled from standard components (Fast-SCNN, self-attention, channel attention) and evaluated against external benchmarks; the claimed contributions (dual-gas fusion, multi-task objective) are not derived from the results they are meant to explain. The self-citations to prior OGI work from the same group (Gasformer [32], GasTwinFormer [35], CarboFormer [14], Embaby et al. [6]) appear as baselines and protocol references, and these prior works do not contain or determine the target pH labels; therefore they are not load-bearing in a circular sense. The remaining concerns—possible session-level visual shortcuts, modality-mask leakage, and the in vitro/in vivo gap—are validity/correctness risks, not cases where a 'prediction' reduces by construction to its inputs. Accordingly, no specific circular step can be quoted.

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

The model's predictions depend on several domain assumptions that are not independently verified in the paper: the pH-gas relationship, the fidelity of OGI cameras to gas concentration, the validity of the in vitro model, and annotation correctness. The hyperparameters listed are hand-chosen but conventional; none of them is fitted to make the target results collapse.

free parameters (4)
  • multi-task loss weight λ = 0.5
    Set by hand in Eq. (8); balances segmentation and classification losses.
  • channel attention reduction ratio = 16
    Architecture hyperparameter for bottleneck MLP in Eq. (5); chosen by hand.
  • self-attention head dimension d_k = 16 (C/8)
    Single-head attention dimension in Eq. (4); fixed by design.
  • segmentation loss weighting = 0.5·focal + 0.5·dice
    Hand-chosen combination for class imbalance and boundary precision.
assumptions (4)
  • domain assumption Under acidosis, rumen CO2 production increases while CH4 production decreases
    Used in Section 1 to motivate dual-gas modality selection; cited to refs [15,23] but treated as established physiology.
  • domain assumption Optical gas imaging (FLIR GF343, GF320) images of gas plumes provide visual signatures that vary with gas concentration/emission rate
    Core to the whole approach; relies on camera manufacturer specs and prior single-gas work; not independently verified in this paper.
  • domain assumption The in vitro ANKOM fermentation system with controlled pH levels is a valid model of rumen gas emissions
    Section 4.1; the ground-truth pH is imposed externally, so the model learns pH-to-image associations in a simplified setting.
  • domain assumption Manual pixel-level annotations of tube, background, and gas plumes are correct
    Training depends on these masks; no inter-annotator agreement is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FUME: Fused Unified Multi-Gas Emission Network for Livestock Rumen Acidosis Detection." pith.science (2026). https://pith.science/paper/54WU5G4A

@misc{pith2026260108205,
  author       = {Pith},
  title        = {Pith review of: FUME: Fused Unified Multi-Gas Emission Network for Livestock Rumen Acidosis Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/54WU5G4A}},
  note         = {Machine review of arXiv:2601.08205}
}
read the original abstract

Ruminal acidosis is a prevalent metabolic disorder in dairy cattle causing significant economic losses and animal welfare concerns. Current diagnostic methods rely on invasive pH measurement, limiting scalability for continuous monitoring. We present FUME (Fused Unified Multi-gas Emission Network), the first deep learning approach for rumen acidosis detection from dual-gas optical imaging under in vitro conditions. Our method leverages complementary carbon dioxide (CO2) and methane (CH4) emission patterns captured by infrared cameras to classify rumen health into Healthy, Transitional, and Acidotic states. FUME employs a lightweight dual-stream architecture with weight-shared encoders, modality-specific self-attention, and channel attention fusion, jointly optimizing gas plume segmentation and classification of dairy cattle health. We introduce the first dual-gas OGI dataset comprising 8,967 annotated frames across six pH levels with pixel-level segmentation masks. Experiments demonstrate that FUME achieves 80.99% mIoU and 98.82% classification accuracy while using only 1.28M parameters and 1.97G MACs--outperforming state-of-the-art methods in segmentation quality with 10x lower computational cost. Ablation studies reveal that CO2 provides the primary discriminative signal and dual-task learning is essential for optimal performance. Our work establishes the feasibility of gas emission-based livestock health monitoring, paving the way for practical, in vitro acidosis detection systems. Codes are available at https://github.com/taminulislam/fume.

Figures

Figures reproduced from arXiv: 2601.08205 by the authors.

Figure 1
Figure 1. Efficiency vs. segmentation quality trade-off across all [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FUME Architecture. Our dual-stream network processes paired CO2 and CH4 thermal images through a weight-shared Fast￾SCNN encoder with depthwise separable convolutions. The Learning-to-Downsample module reduces spatial resolution to H 8 × W 8 , while the Global Feature Extractor with Pyramid Pooling Module captures multi-scale context at H 32 × W 32 . Modality-specific self-attention modules refine each stream before… view at source ↗
Figure 3
Figure 3. Comparison with state-of-the-art methods. (A) Mean IoU comparison showing FUME substantially outperforms both general [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Qualitative segmentation results across Acidotic, Transitional, and Healthy classes for both CO [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Ablation study results. (A) Dice coefficient comparison across architectural variants, demonstrating that the proposed FUME [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 5 linked inside Pith

  1. [1]

    Compari- son of manual and automated body condition scoring of com- mercial dairy cattle.Veterinary Record, 195(7):no–no, 2024

    Tom Robert Angel and Sophie Anne Mahendran. Compari- son of manual and automated body condition scoring of com- mercial dairy cattle.Veterinary Record, 195(7):no–no, 2024. 3

    2024

  2. [2]

    Progressively complementarity- aware fusion network for rgb-d salient object detection

    Hao Chen and Youfu Li. Progressively complementarity- aware fusion network for rgb-d salient object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3051–3060, 2018. 3

    2018

  3. [3]

    Adaptive fusion of single-view and multi-view depth for autonomous driving

    Junda Cheng, Wei Yin, Kaixuan Wang, Xiaozhi Chen, Shijie Wang, and Xin Yang. Adaptive fusion of single-view and multi-view depth for autonomous driving. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10138–10147, 2024. 3

    2024

  4. [4]

    Subacute ruminal acidosis in cattle: A critical review of clinical management.Veterinary Research Communications, 49(5):273, 2025

    Georgios Christodoulopoulos. Subacute ruminal acidosis in cattle: A critical review of clinical management.Veterinary Research Communications, 49(5):273, 2025. 1

    2025

  5. [5]

    Effect of ru- men culture adaptation to origanum vulgare l

    MG Embaby, M G ¨unal, and A AbuGhazaleh. Effect of ru- men culture adaptation to origanum vulgare l. essential oil on rumen methane and fermentation.American J. Anim. Vet. Sci, 14:190–196, 2019. 5

    2019

  6. [6]

    Optical gas imaging and deep learning for quantifying enteric methane emissions from rumen fermentation in vitro.IET Image Processing, 19(1):e13327, 2025

    Mohamed G Embaby, Toqi Tahamid Sarker, Amer AbuG- hazaleh, and Khaled R Ahmed. Optical gas imaging and deep learning for quantifying enteric methane emissions from rumen fermentation in vitro.IET Image Processing, 19(1):e13327, 2025. 2

    2025

  7. [7]

    The monitoring, prevention and treat- ment of sub-acute ruminal acidosis (sara): A review.The veterinary journal, 176(1):32–43, 2008

    J ¨org MD Enemark. The monitoring, prevention and treat- ment of sub-acute ruminal acidosis (sara): A review.The veterinary journal, 176(1):32–43, 2008. 1

    2008

  8. [8]

    Dual attention network for scene seg- mentation

    Jun Fu, Jing Liu, Haijie Tian, Yong Li, Yongjun Bao, Zhiwei Fang, and Hanqing Lu. Dual attention network for scene seg- mentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3146–3154,

Show all 46 references
  1. [9]

    Cafnet: Cross-modal adaptive fusion network with attention and gated weighting for rgb-t semantic segmentation.IEEE Access, 2025

    Meili Fu, Huanliang Sun, Zhihan Chen, and Lulin Wei. Cafnet: Cross-modal adaptive fusion network with attention and gated weighting for rgb-t semantic segmentation.IEEE Access, 2025. 2

    2025

  2. [10]

    Prevalence, persistence, and costs of acidosis in holstein cows across breeding regions

    Vesna Gantner, Franjo Poljak, Boris Ljubojevi ´c, Zvonimir Steiner, and Ranko Gantner. Prevalence, persistence, and costs of acidosis in holstein cows across breeding regions. AgroReS, 14:158–165, 2025. 1

    2025

  3. [11]

    Deep dual-resolution networks for real-time and accu- rate semantic segmentation of road scenes.arXiv preprint arXiv:2101.06085, 2021

    Yuanduo Hong, Huihui Pan, Weichao Sun, and Yisong Jia. Deep dual-resolution networks for real-time and accu- rate semantic segmentation of road scenes.arXiv preprint arXiv:2101.06085, 2021. 4, 5, 6

    2021 arXiv

  4. [12]

    A dual-stream feature decomposition network with weight transformation for multi-modality image fusion

    Tianqing Hu, Xiaofei Nan, Xiabing Zhou, Yu Shen, and Qin- glei Zhou. A dual-stream feature decomposition network with weight transformation for multi-modality image fusion. Scientific Reports, 15(1):7467, 2025. 3

    2025

  5. [13]

    Weedswin hierarchical vision transformer with sam-2 for multi-stage weed detection and classification.Scientific Reports, 15(1): 23274, 2025

    Taminul Islam, Toqi Tahamid Sarker, Khaled R Ahmed, Cristiana Bernardi Rankrape, and Karla Gage. Weedswin hierarchical vision transformer with sam-2 for multi-stage weed detection and classification.Scientific Reports, 15(1): 23274, 2025. 3

    2025

  6. [14]

    Taminul Islam, Toqi Tahamid Sarker, Mohamed G Embaby, Khaled R Ahmed, and Amer AbuGhazaleh. Carbonext and carboformer: Dual semantic segmentation architectures for detecting and quantifying carbon dioxide emissions using optical gas imaging.arXiv preprint arXiv:2506.05360, 202...

    2025 arXiv

  7. [15]

    Peter H Janssen. Influence of hydrogen on rumen methane formation and fermentation balances through microbial growth kinetics and fermentation thermodynamics.Animal Feed Science and Technology, 160(1-2):1–22, 2010. 1

    2010

  8. [16]

    Vision-based real-time marine and off- shore structural health monitoring system using underwater robots.Computer-Aided Civil and Infrastructure Engineer- ing, 39(2):281–299, 2024

    Pengcheng Jiao, Xinghong Ye, Chenjie Zhang, Wentao Li, and Haipeng Wang. Vision-based real-time marine and off- shore structural health monitoring system using underwater robots.Computer-Aided Civil and Infrastructure Engineer- ing, 39(2):281–299, 2024. 3

    2024

  9. [17]

    Evolution of multiorgan segmentation techniques from traditional to deep learning in abdominal ct images–a systematic review.Dis- plays, 73:102223, 2022

    Harinder Kaur, Navjot Kaur, and Nirvair Neeru. Evolution of multiorgan segmentation techniques from traditional to deep learning in abdominal ct images–a systematic review.Dis- plays, 73:102223, 2022. 6

    2022

  10. [18]

    Cross- modal learning with 3d deformable attention for action recognition

    Sangwon Kim, Dasom Ahn, and Byoung Chul Ko. Cross- modal learning with 3d deformable attention for action recognition. InProceedings of the IEEE/CVF international conference on computer vision, pages 10265–10275, 2023. 3

    2023

  11. [19]

    Ssm-net: Semi-supervised multi-task network for joint le- sion segmentation and classification from pancreatic eus im- ages.Artificial Intelligence in Medicine, 154:102919, 2024

    Jiajia Li, Pingping Zhang, Xia Yang, Lei Zhu, Teng Wang, Ping Zhang, Ruhan Liu, Bin Sheng, and Kaixuan Wang. Ssm-net: Semi-supervised multi-task network for joint le- sion segmentation and classification from pancreatic eus im- ages.Artificial Intelligence in Medicine, 154:102...

    2024

  12. [20]

    A novel lameness detection method for dairy cows based on temporal gait and spatial post features.Com- puters and Electronics in Agriculture, 224:109172, 2024

    Qian Li, Yongsheng Si, Mengyuan Chu, Ning Liu, Xi Kang, and Gang Liu. A novel lameness detection method for dairy cows based on temporal gait and spatial post features.Com- puters and Electronics in Agriculture, 224:109172, 2024. 3

    2024

  13. [21]

    Mmtl-uniad: A unified framework for multimodal and multi-task learning in assistive driving perception

    Wenzhuo Liu, Wenshuo Wang, Yicheng Qiao, Qiannan Guo, Jiayin Zhu, Pengfei Li, Zilong Chen, Huiming Yang, Zhiwei Li, Lening Wang, et al. Mmtl-uniad: A unified framework for multimodal and multi-task learning in assistive driving perception. InProceedings of the Computer Vision ...

    2025

  14. [22]

    Vehicle–canine collaboration for urban pipeline methane leak detection.Nature Cities, 2(4):336– 343, 2025

    Hongfang Lu, Dongmin Xi, Yaqin Xiang, Zhenhao Su, and Y Frank Cheng. Vehicle–canine collaboration for urban pipeline methane leak detection.Nature Cities, 2(4):336– 343, 2025. 2

    2025

  15. [23]

    Rumen acidosis in ruminants: a review of the effects of high-concentrate diets and the poten- tial modulatory role of rumen foam.Frontiers in Veterinary Science, 12:1595615, 2025

    Jinju Mao and Lizhi Wang. Rumen acidosis in ruminants: a review of the effects of high-concentrate diets and the poten- tial modulatory role of rumen foam.Frontiers in Veterinary Science, 12:1595615, 2025. 1

    2025

  16. [24]

    Espnetv2: A light-weight, power ef- ficient, and general purpose convolutional neural network

    Sachin Mehta, Mohammad Rastegari, Linda Shapiro, and Hannaneh Hajishirzi. Espnetv2: A light-weight, power ef- ficient, and general purpose convolutional neural network. InProceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 9190–9200, 2019...

    2019

  17. [25]

    Revolutionizing crop dis- ease detection with computational deep learning: a compre- hensive review.Environmental Monitoring and Assessment, 196(3):302, 2024

    Habiba N Ngugi, Absalom E Ezugwu, Andronicus A Akinyelu, and Laith Abualigah. Revolutionizing crop dis- ease detection with computational deep learning: a compre- hensive review.Environmental Monitoring and Assessment, 196(3):302, 2024. 3

    2024

  18. [26]

    Acidosis in cattle: a review.Journal of animal science, 76(1):275–286,

    FN Owens, DS Secrist, WJ Hill, and DR Gill. Acidosis in cattle: a review.Journal of animal science, 76(1):275–286,

  19. [27]

    Enet: A deep neural network architec- ture for real-time semantic segmentation.arXiv preprint arXiv:1606.02147, 2016

    Adam Paszke, Abhishek Chaurasia, Sangpil Kim, and Eu- genio Culurciello. Enet: A deep neural network architec- ture for real-time semantic segmentation.arXiv preprint arXiv:1606.02147, 2016. 4, 5, 6

    2016 arXiv

  20. [28]

    Sub- acute ruminal acidosis in dairy cows: The physiological causes, incidence and consequences.The Veterinary Jour- nal, 176(1):21–31, 2008

    JC Plaizier, DO Krause, GN Gozho, and BW McBride. Sub- acute ruminal acidosis in dairy cows: The physiological causes, incidence and consequences.The Veterinary Jour- nal, 176(1):21–31, 2008. 1, 6

    2008

  21. [29]

    Fast-scnn: Fast semantic segmentation network.arXiv preprint arXiv:1902.04502, 2019

    Rudra PK Poudel, Stephan Liwicki, and Roberto Cipolla. Fast-scnn: Fast semantic segmentation network.arXiv preprint arXiv:1902.04502, 2019. 4

    1902 arXiv

  22. [30]

    Residual transformer fusion network for salt and pepper image denoising.arXiv preprint arXiv:2502.09000,

    Bintang Pradana Erlangga Putra, Heri Prasetyo, and Esti Suryani. Residual transformer fusion network for salt and pepper image denoising.arXiv preprint arXiv:2502.09000,

    arXiv

  23. [31]

    La-mtl: Latency- aware automated multi-task learning

    Shambhavi Balamuthu Sampath, Sami Sawani, Moritz Thoma, Lukas Frickenstein, Pierpaolo Mori, Nael Fas- fous, Manoj Rohit Vemparala, Alexander Frickenstein, Ulf Schlichtmann, Claudio Passerone, et al. La-mtl: Latency- aware automated multi-task learning. In2025 62nd ACM/IEEE Des...

  24. [32]

    Gasformer: A transformer-based architecture for segmenting methane emissions from livestock in optical gas imaging

    Toqi Tahamid Sarker, Mohamed G Embaby, Khaled R Ahmed, and Amer AbuGhazaleh. Gasformer: A transformer-based architecture for segmenting methane emissions from livestock in optical gas imaging. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

    2024

  25. [33]

    Weedsense: Multi-task learning for weed segmentation, height estima- tion, and growth stage classification

    Toqi Tahamid Sarker, Khaled R Ahmed, Taminul Islam, Cristiana Bernardi Rankrape, and Karla Gage. Weedsense: Multi-task learning for weed segmentation, height estima- tion, and growth stage classification. InProceedings of the IEEE/CVF International Conference on Computer Visio...

    2025

  26. [34]

    Fusing satellite im- agery and ground-based observations for pm2

    Zohreh Sohrabi and Jamshid Maleki. Fusing satellite im- agery and ground-based observations for pm2. 5 air pollution modeling in iran using a deep learning approach.Scientific Reports, 15(1):21449, 2025. 2

    2025

  27. [35]

    Gastwinformer: A hybrid vision transformer for livestock methane emission segmentation and dietary classification in optical gas imag- ing.arXiv e-prints, pages arXiv–2508, 2025

    Toqi Tahamid Sarker, Mohamed Embaby, Taminul Islam, Amer AbuGhazaleh, and Khaled R Ahmed. Gastwinformer: A hybrid vision transformer for livestock methane emission segmentation and dietary classification in optical gas imag- ing.arXiv e-prints, pages arXiv–2508, 2025. 2, 4, 5, 6, 7

    2025

  28. [36]

    Swinlstm: Improving spatiotemporal prediction accuracy using swin transformer and lstm

    Song Tang, Chuang Li, Pu Zhang, and RongNian Tang. Swinlstm: Improving spatiotemporal prediction accuracy using swin transformer and lstm. InProceedings of the IEEE/CVF international conference on computer vision, pages 13470–13479, 2023. 3

    2023

  29. [37]

    Mti-net: Multi-scale task interaction networks for multi-task learning

    Simon Vandenhende, Stamatios Georgoulis, and Luc Van Gool. Mti-net: Multi-scale task interaction networks for multi-task learning. InEuropean conference on computer vision, pages 527–543. Springer, 2020. 4, 5, 6

    2020

  30. [38]

    Nikolaos V oulgarakis, Dimitrios Gougoulis, Dimitra Psalla, Georgios Papakonstantinou, Panagiotis-Dimitrios Katsou- los, Konstantinos Katsoulis, Mariana Angelidou-Tsifida, Labrini Athanasiou, Vasilleios Papatsiros, and Georgios Christodoulopoulos. Can computerized rumen mucosa...

    2024

  31. [39]

    Subacute rumen acidosis in greek dairy sheep: Prevalence, impact and colorimetry management.Animals, 14(14):2061,

    Nikolaos V oulgarakis, Dimitrios A Gougoulis, Dimitra Psalla, Georgios I Papakonstantinou, Konstantinos Kat- soulis, Mariana Angelidou-Tsifida, Labrini V Athanasiou, Vasileios G Papatsiros, and Georgios Christodoulopoulos. Subacute rumen acidosis in greek dairy sheep: Prevalen...

  32. [40]

    Guanbo Wang, Haiyan Li, Shuhua Ye, Hongzhi Zhao, Hong- wei Ding, and Shidong Xie. Rfwnet: A multiscale remote sensing forest wildfire detection network with digital twin- ning, adaptive spatial aggregation, and dynamic sparse fea- tures.IEEE Transactions on Geoscience and Remo...

    2024

  33. [41]

    Hierarchical multi-modal fusion for roadside vru detection: Method complementarity under sparse label constraints

    Chuheng Wei, Ziyan Zhang, Haishan Liu, Guoyuan Wu, and Matthew Barth. Hierarchical multi-modal fusion for roadside vru detection: Method complementarity under sparse label constraints. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 5343–5350, 2025. 3

    2025

  34. [42]

    Openemma: Open-source multimodal model for end-to-end autonomous driving

    Shuo Xing, Chengyuan Qian, Yuping Wang, Hongyuan Hua, Kexin Tian, Yang Zhou, and Zhengzhong Tu. Openemma: Open-source multimodal model for end-to-end autonomous driving. InProceedings of the Winter Conference on Appli- cations of Computer Vision, pages 1001–1009, 2025. 3

    2025

  35. [43]

    A unified framework for 3d scene understanding.Advances in Neural Information Processing Systems, 37:59468–59490, 2024

    Wei Xu, Chunsheng Shi, Sifan Tu, Xin Zhou, Dingkang Liang, and Xiang Bai. A unified framework for 3d scene understanding.Advances in Neural Information Processing Systems, 37:59468–59490, 2024. 3

    2024

  36. [44]

    Bisenet v2: Bilateral net- work with guided aggregation for real-time semantic seg- mentation.International journal of computer vision, 129 (11):3051–3068, 2021

    Changqian Yu, Changxin Gao, Jingbo Wang, Gang Yu, Chunhua Shen, and Nong Sang. Bisenet v2: Bilateral net- work with guided aggregation for real-time semantic seg- mentation.International journal of computer vision, 129 (11):3051–3068, 2021. 4, 5, 6, 7

    2021

  37. [45]

    Cmx: Cross-modal fusion for rgb-x semantic segmentation with transformers.IEEE Transactions on intelligent transportation systems, 24(12): 14679–14694, 2023

    Jiaming Zhang, Huayao Liu, Kailun Yang, Xinxin Hu, Ruip- ing Liu, and Rainer Stiefelhagen. Cmx: Cross-modal fusion for rgb-x semantic segmentation with transformers.IEEE Transactions on intelligent transportation systems, 24(12): 14679–14694, 2023. 4, 5, 6

    2023

  38. [46]

    A time- series neural network for pig feeding behavior recognition and dangerous detection from videos.Computers and Elec- tronics in Agriculture, 218:108710, 2024

    Yan Zhang, Xinze Yang, Yufei Liu, Junyu Zhou, Yihong Huang, Jiapeng Li, Longxiang Zhang, and Qin Ma. A time- series neural network for pig feeding behavior recognition and dangerous detection from videos.Computers and Elec- tronics in Agriculture, 218:108710, 2024. 3

    2024

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.