Please enable JavaScript.

Coggle requires JavaScript to display documents.

CNN detects birds' sound - Coggle Diagram

- - - - AlexNet
        VGG-19
        Inception v3
        RestNet-34
        ResNet-152
        DenseNet-169
      - Spectrogram
        frames: 46.4ms
        2048bings, 44.1kHz
        50% overlap
    - - VGG-16
        ResNet-50
      - Spectrogram
        clip: 2s
        sampling rate: 48kHz
        NFFT: 2048
        hop length: 512
        window length: 2048
        hann window
    - - Discrete Gabor Transform (DGT)
      - GoogleNet
        VGG-16
        VGG-19
    - - Spectrogram
        Discrete Fourier Transform(DFT)
        1024 samples
      - Harmonic and percussion images
        Harmonic-Percussive Sound Separation (HPSS)
      - Scattergram
      - AlexNet
        GoogleNet
        VGG-16
        VGG-19
        ResNet-50
        Inception v3
    - - Mel-frequency cepstral coefficients (MFCCs)
      - self designed CNN architecture
        (VGG styled)
    - - Mel-Spectrogram
      - ResNet-50
    - - mel-spectrogram
        Per-channel energy normalization(PCEN)
      - self designed model
        Context-adaptive neural network layer
    - - Spectrogram
      - Mel-spectrogram
        2048 samples
    - - Spectrogram
      - self designed model
    - - mel-spectrogram
      - VGG style network
    - - MobileNetV2
    - - Three self designed architecture
    - - Siamese Neural Network (SNN) (VGG style)
        GoogleNet
        VGG-16
        VGG-19
        GoogleNetP365
        eCNN
    - - self designed network
      - spectrogram
    - - Residual nets
      - Standard CNNs
      - Densely connected nets
    - - Hyperparameters
      - Equations
    - - Mel-scaled STFT
      - Shallower CNN models
      - Linear STFT
    - - Hyperparameters
    - - YOLOv3
      - two-stage feature extraction
    - - ANN
      - AEM
      - PPFM
    - - on multi-channel
      - DCNN
        VGG-16
    - - Unlabeled video
  - - - log spectrum
        log mel-band energies
      - CNN
        CRNN
  - - - Mel-Spectrogram
      - ResNet-50
      - Frog
    - - R-CNN
        SSD
        YOLO
        R-FCN
        UAV
    - - AUC-ROC
    - - pre-trained alexnet
      - high spatial resolution
    - - Pneumonia