Fish image recognition method based on multi-layer feature fusion convolutional network

Li, Lipeng; Shi, Feipeng; Wang, Chunxu

doi:10.1016/j.ecoinf.2022.101873

Cited by 9 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Simi-larly, Villon et al [14] used a convolutional neural network to analyze images from social media, providing support in monitoring rare megafauna species. Li et al [15] proposed Tripmix-Net, a fish image classification model that incorporates multiscale network fusion. Qu et al [16] introduced DAMNet, a deep neural network with a dual-attention mechanism for aquatic biological image classification.…”

Section: Introductionmentioning

confidence: 99%

DiffusionFR: Species Recognition of Fish in Blurry Scenarios via Diffusion and Attention

Wang,

Shi,

et al. 2024

Animals

View full text Add to dashboard Cite

Blurry scenarios, such as light reflections and water ripples, often affect the clarity and signal-to-noise ratio of fish images, posing significant challenges for traditional deep learning models in accurately recognizing fish species. Firstly, deep learning models rely on a large amount of labeled data. However, it is often difficult to label data in blurry scenarios. Secondly, existing deep learning models need to be more effective for the processing of bad, blurry, and otherwise inadequate images, which is an essential reason for their low recognition rate. A method based on the diffusion model and attention mechanism for fish image recognition in blurry scenarios, DiffusionFR, is proposed to solve these problems and improve the performance of species recognition of fish images in blurry scenarios. This paper presents the selection and application of this correcting technique. In the method, DiffusionFR, a two-stage diffusion network model, TSD, is designed to deblur bad, blurry, and otherwise inadequate fish scene pictures to restore clarity, and a learnable attention module, LAM, is intended to improve the accuracy of fish recognition. In addition, a new dataset of fish images in blurry scenarios, BlurryFish, was constructed and used to validate the effectiveness of DiffusionFR, combining bad, blurry, and otherwise inadequate images from the publicly available dataset Fish4Knowledge. The experimental results demonstrate that DiffusionFR achieves outstanding performance on various datasets. On the original dataset, DiffusionFR achieved the highest training accuracy of 97.55%, as well as a Top-1 accuracy test score of 92.02% and a Top-5 accuracy test score of 95.17%. Furthermore, on nine datasets with light reflection noise, the mean values of training accuracy reached a peak at 96.50%, while the mean values of the Top-1 accuracy test and Top-5 accuracy test were at their highest at 90.96% and 94.12%, respectively. Similarly, on three datasets with water ripple noise, the mean values of training accuracy reached a peak at 95.00%, while the mean values of the Top-1 accuracy test and Top-5 accuracy test were at their highest at 89.54% and 92.73%, respectively. These results demonstrate that the method showcases superior accuracy and enhanced robustness in handling original datasets and datasets with light reflection and water ripple noise.

show abstract

Section: Introductionmentioning

confidence: 99%

DiffusionFR: Species Recognition of Fish in Blurry Scenarios via Diffusion and Attention

Wang,

Shi,

et al. 2024

Animals

View full text Add to dashboard Cite

show abstract

“…With China's rapidly developing power transmission and distribution grid [1], the application of distributed generation and nonlinear load in the power grid is becoming more and more extensive. For example, access to a large number of solid-state switches, power electronic switches, nonlinear loads will cause serious distortion of grid voltage and current, and the deterioration of power quality will also increase, which will affect the stability of the power grid equipment [2]. Therefore, the power quality disturbances in line identification and governance are of great significance.…”

Section: Introductionmentioning

confidence: 99%

Power Quality Disturbance Identification Basing on Adaptive Kalman Filter and Multi-Scale Channel Attention Fusion Convolutional Network

Zhao,

Liu,

Chen

et al. 2024

ENERGY

View full text Add to dashboard Cite

In light of the prevailing issue that the existing convolutional neural network (CNN) power quality disturbance identification method can only extract single-scale features, which leads to a lack of feature information and weak anti-noise performance, a new approach for identifying power quality disturbances based on an adaptive Kalman filter (KF) and multi-scale channel attention (MS-CAM) fused convolutional neural network is suggested. Single and composite-disruption signals are generated through simulation. The adaptive maximum likelihood Kalman filter is employed for noise reduction in the initial disturbance signal, and subsequent integration of multi-scale features into the conventional CNN architecture is conducted. The multi-scale features of the signal are captured by convolution kernels of different sizes so that the model can obtain diverse feature expressions. The attention mechanism (ATT) is introduced to adaptively allocate the extracted features, and the features are fused and selected to obtain the new main features. The Softmax classifier is employed for the classification of power quality disturbances. Finally, by comparing the recognition accuracy of the convolutional neural network (CNN), the model using the attention mechanism, the bidirectional long-term and short-term memory network (MS-Bi-LSTM), and the multi-scale convolutional neural network (MSCNN) with the attention mechanism with the proposed method. The simulation results demonstrate that the proposed method is higher than CNN, MS-Bi-LSTM, and MSCNN, and the overall recognition rate exceeds 99%, and the proposed method has significant classification accuracy and robust classification performance. This achievement provides a new perspective for further exploration in the field of power quality disturbance classification.

show abstract

Deep Fish: An Approach to Fish Species Identification Through Deep Learning Techniques

Reddy,

Malathi,

Julaiha

2024

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

Fish image recognition method based on multi-layer feature fusion convolutional network

Cited by 9 publications

References 13 publications

DiffusionFR: Species Recognition of Fish in Blurry Scenarios via Diffusion and Attention

DiffusionFR: Species Recognition of Fish in Blurry Scenarios via Diffusion and Attention

Power Quality Disturbance Identification Basing on Adaptive Kalman Filter and Multi-Scale Channel Attention Fusion Convolutional Network

Deep Fish: An Approach to Fish Species Identification Through Deep Learning Techniques

Contact Info

Product

Resources

About