The project aims to use Convolutional Neural Networks (CNN) to identify mosquito sounds.
Data Collection: HumBugDB: A large-scale acoustic mosquito dataset.
Spectrogram Conversion:The sound of mosquitoes has distinct frequency characteristics.
Figure 1: spectrogram
Figure 2: Architecture
To address the data sample imbalance, two different methods were employed to train the model.
Model 1: Trained on small datasets using 6-fold cross-validation.
Model 2: Trained on large-scale data, with adjustments made to the weights of different categories.
Figure 3: model loss and performance
Model 1 has weaker performance but trains very quickly, whereas Model 2 has slower training speed but better overall performance, particularly in mosquito sound detection.
There are several tips to speed up model training:

By applying these methods, the model can converge within one minute on a small dataset of 1800 images and within approximately five minutes on a large-scale dataset of 30000 images.
To further enhance the performance and accuracy of the model, we will implement a series of optimization measures.