Deep residual convolutional neural network based on hybrid attention mechanism for ecological monitoring of marine fishery

Liu, J.; Zhang, Lei; Li, Yanfei; Liu, Hui

doi:10.1016/j.ecoinf.2023.102204

Cited by 13 publications

(3 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, three distinct Gaussian noise levels, 0.01, 0.05, and 0.1, were introduced to the noise-free images, replicating the realistic thermographic scenarios [33]. The mathematical expression of the Gaussian noise applied to thermal images is formulated in equation ( 2) [34]. This inclusion of noisy data allows for assessing the model's generalizability under varying conditions.…”

Section: Data Collectionmentioning

confidence: 99%

CNN-Based Approach for Non-Invasive Estimation of Breast Tumor Size and Location Using Thermographic Images

Khomsi,

El Fezazi,

Bellarbi

2024

Int. J. Onl. Eng.

View full text Add to dashboard Cite

The characterization of tumors is crucial for guiding appropriate treatment strategies and enhancing patient survival rates. Surface thermography shows promise in the non-invasive detection of thermal patterns associated with the existence of breast tumors. Nevertheless, the precise prediction of both tumor size and location using temperature characteristics presents a critical challenge. This is due to the limited availability of thermal images labeled with the corresponding tumor size and location. This work proposes a deep learning approach based on convolutional neural networks (CNN) in combination with thermographic images for estimating breast tumor size and location. Successive COMSOL-based simulations are conducted, including a 3D breast model with various tumor scenarios. Thus, different noise levels were included in the development of the thermographic image dataset. Every image was accordingly labeled with the corresponding tumor location and size to train the CNN model. Mean absolute error (MAE) and the coefficient of determination (R²) were considered as evaluation metrics. The results show that the proposed CNN model achieved a reasonable prediction performance with MAE–R² values of 0.872–98.6% for tumor size, 1.161–96.8% for x location, 1.086–97.1% for y location, and 0.954–96.7% for z location. This study indicates that the combination of surface thermography and deep learning is a convenient tool for predicting breast tumor parameters.

show abstract

Section: Data Collectionmentioning

confidence: 99%

CNN-Based Approach for Non-Invasive Estimation of Breast Tumor Size and Location Using Thermographic Images

Khomsi,

El Fezazi,

Bellarbi

2024

Int. J. Onl. Eng.

View full text Add to dashboard Cite

show abstract

“…In To improve the regression accuracy of the LSTM and BiGRU, we also added mechanisms of attention to the structure of the deep learning network [45]. Suppose the input vectors are the multidimensional feature vectors before the predicted time.…”

Section: Deep Learning Modelsmentioning

confidence: 99%

“…Suppose the input vectors are the multidimensional feature vectors before the predicted time. The attention coefficient is obtained by calculating the previously hidden layer To improve the regression accuracy of the LSTM and BiGRU, we also added mechanisms of attention to the structure of the deep learning network [45]. Suppose the input vectors are the multidimensional feature vectors before the predicted time.…”

Section: Deep Learning Modelsmentioning

confidence: 99%

Wearable Bioimpedance-Based Deep Learning Techniques for Live Fish Health Assessment under Waterless and Low-Temperature Conditions

Zhang,

Chen,

Feng

et al. 2023

Sensors

View full text Add to dashboard Cite

(1) Background: At present, physiological stress detection technology is a critical means for precisely evaluating the comprehensive health status of live fish. However, the commonly used biochemical tests are invasive and time-consuming and cannot simultaneously monitor and dynamically evaluate multiple stress levels in fish and accurately classify their health levels. The purpose of this study is to deploy wearable bioelectrical impedance analysis (WBIA) sensors on fish skin to construct a deep learning-based stress dynamic evaluation model for precisely estimating their accurate health status. (2) Methods: The correlation of fish (turbot) muscle nutrients and their stress indicators are calculated using grey relation analysis (GRA) for allocating the weight of the stress factors. Next, WBIA features are sieved using the maximum information coefficient (MIC) in stress trend evaluation modeling, which is closely related to the key stress factors. Afterward, a convolutional neural network (CNN) is utilized to obtain the features of the WBIA signals. Then, the long short-term memory (LSTM) method learns the stress trends with residual rectification using bidirectional gated recurrent units (BiGRUs). Furthermore, the Z-shaped fuzzy function can accurately classify the fish health status by the total evaluated stress values. (3) Results: The proposed CNN-LSTM-BiGRU-based stress evaluation model shows superior accuracy compared to the other machine learning models (CNN-LSTM, CNN-GRU, LSTM, GRU, SVR, and BP) based on the MAPE, MAE, and RMSE. Moreover, the fish health classification under waterless and low-temperature conditions is thoroughly verified. High accuracy is proven by the classification validation criterion (accuracy, F1 score, precision, and recall). (4) Conclusions: the proposed health evaluation technology can precisely monitor and track the health status of live fish and provides an effective technical reference for the field of live fish vital sign detection.

show abstract

Gaussian mutation-alpine skiing optimization algorithm-recurrent attention unit-gated recurrent unit-extreme learning machine model: an advanced predictive model for predicting evaporation

Ehteram,

Barzegari Banadkooki,

Afshari Nia

2024

Stoch Environ Res Risk Assess

View full text Add to dashboard Cite

Deep residual convolutional neural network based on hybrid attention mechanism for ecological monitoring of marine fishery

Cited by 13 publications

References 56 publications

CNN-Based Approach for Non-Invasive Estimation of Breast Tumor Size and Location Using Thermographic Images

CNN-Based Approach for Non-Invasive Estimation of Breast Tumor Size and Location Using Thermographic Images

Wearable Bioimpedance-Based Deep Learning Techniques for Live Fish Health Assessment under Waterless and Low-Temperature Conditions

Gaussian mutation-alpine skiing optimization algorithm-recurrent attention unit-gated recurrent unit-extreme learning machine model: an advanced predictive model for predicting evaporation

Contact Info

Product

Resources

About