Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Experimental system diagram.

More »

Fig 1 Expand

Fig 2.

RGB images of three different feeding states.

(A) None Feeding State. (B) Weak Feeding State. (C) Strong Feeding State.

More »

Fig 2 Expand

Table 1.

Classification criteria of feeding state intensity.

More »

Table 1 Expand

Fig 3.

Model structure diagram.

More »

Fig 3 Expand

Fig 4.

Optical flow images of three different feeding states.

(A) None Feeding State. (B) Weak Feeding State. (C) Strong Feeding State.

More »

Fig 4 Expand

Fig 5.

The basic structure of ResNet-50.

More »

Fig 5 Expand

Fig 6.

ResNet network residual unit structure.

More »

Fig 6 Expand

Fig 7.

Binary images of three different feeding states.

(A) None Feeding State. (B) Weak Feeding State. (C) Strong Feeding State.

More »

Fig 7 Expand

Fig 8.

Basic structure of the one-dimensional convolutional neural network.

More »

Fig 8 Expand

Fig 9.

The training and testing results for the temporal feature network (A), spatial feature network (B), and data statistical feature network (C).

More »

Fig 9 Expand

Table 2.

Network training results.

More »

Table 2 Expand

Fig 10.

Three confusion matrix diagrams depicting the test results based on temporal feature network (A), spatial feature network (B), and data statistical feature network (C).

More »

Fig 10 Expand

Table 3.

Evaluation index of classification results of fish feeding status by three-stream network.

More »

Table 3 Expand

Table 4.

Ablation study.

More »

Table 4 Expand

Table 5.

Comparison with other models for fish feeding behavior discrimination.

More »

Table 5 Expand

Fig 11.

Class activation mapping (CAM) of test samples.

(A), (B), and (C) are samples with correct predictions, classed as None, Weak, and Strong. (D) denotes the misclassified sample. Its essential actual value is Weak, and the predicted values are None.

More »

Fig 11 Expand

Table 6.

Comparison of processing time of different feature network models.

More »

Table 6 Expand

Fig 12.

Software function interface.

More »

Fig 12 Expand