U-stage day 8

5418 ワード

1.授業内容

[DLBasic]Convolutionとは?

CNN(Convolutional Neural Networks)
CNN consists of convolution layer, pooling layer, and fully connected layer.
1)Convolution and pooling layers: feature extraction
2)Fully connected layer: decision making(e.g., classification)

[DL Basic]Modern CNN

1)VGG: repeated 3x3 blocks
2)GoogleNet: 1x1 convolution
3)ResNet: skip-connection
4)DenseNet: concatenation

[DL Basic]Computer Vision

Fully Convolutional Network
一般的なCNNの様子

フルボリュームネットワークの様子

Detection
1)R_CNN
1.takes an input image,
2.extracts around 2,000 region proposals (using Selective search),
3.compute features for each proposal (using AlexNet), and then
4.classifies with linear SVMs.

2)SPPNet
1.In R-CNN, the number of crop/warp is usually over 2,000 meaning that CNN must run more than 2,000 times (59s/image on CPU).
2.However, in SPPNet, CNN runs once.
3)Fast R_CNN
1.Takes an input and a set of bounding boxes.
2.Generated convolutional feature map
3.For each region, get a fixed length feature from ROI pooling
4.Two outputs: class and bounding-box regressor.
4)YOLO
1.YOLO (v1) is an extremely fast object detection algorithm
2.It simultaneously predicts multiple bounding boxes and class probabilities
3.Given an image, YOLO divides it into SxS grid
4.Each cell predicts B bounding boxes (B=5)
5.Each cell predicts C class probabilities

実習
必修科目3内容

2.課題実行過程/成果整理

Define MLP model

class ConvolutionalNeuralNetworkClass(nn.Module):
    """
        Convolutional Neural Network (CNN) Class
    """
    def __init__(self,name='cnn',xdim=[1,28,28],
                 ksize=3,cdims=[32,64],hdims=[1024,128],ydim=10,
                 USE_BATCHNORM=False):
        super(ConvolutionalNeuralNetworkClass,self).__init__()
        self.name = name
        self.xdim = xdim
        self.ksize = ksize
        self.cdims = cdims
        self.hdims = hdims
        self.ydim = ydim
        self.USE_BATCHNORM = USE_BATCHNORM

        # Convolutional layers
        self.layers = []
        prev_cdim = self.xdim[0]
        for cdim in self.cdims: # for each hidden layer
            self.layers.append(
                nn.Conv2d(
                    in_channels = prev_cdim,
                    out_channels = cdim,
                    kernel_size = self.ksize,
                    stride = (1, 1),
                    padding = self.ksize//2
                )) # convlution 
            if self.USE_BATCHNORM:
                self.layers.append(nn.BatchNorm2d(cdim)) # batch-norm
            self.layers.append(nn.ReLU(True))  # activation
            self.layers.append(nn.MaxPool2d(kernel_size=(2,2), stride=(2,2))) # max-pooling 
            self.layers.append(nn.Dropout2d(p=0.5))  # dropout
            prev_cdim = cdim

        # Dense layers
        self.layers.append(nn.Flatten())
        prev_hdim = prev_cdim*(self.xdim[1]//(2**len(self.cdims)))*(self.xdim[2]//(2**len(self.cdims)))
        for hdim in self.hdims:
            self.layers.append(nn.Linear(
                prev_hdim, hdim, bias = True
                               ))
            self.layers.append(nn.ReLU(True))  # activation
            prev_hdim = hdim
        # Final layer (without activation)
        self.layers.append(nn.Linear(prev_hdim,self.ydim,bias=True))

        # Concatenate all layers 
        self.net = nn.Sequential()
        for l_idx,layer in enumerate(self.layers):
            layer_name = "%s_%02d"%(type(layer).__name__.lower(),l_idx)
            self.net.add_module(layer_name,layer)
        self.init_param() # initialize parameters
        
    def init_param(self):
        for m in self.modules():
            if isinstance(m,nn.Conv2d): # init conv
                nn.init.kaiming_normal_(m.weight)
                nn.init.zeros_(m.bias)
            elif isinstance(m,nn.BatchNorm2d): # init BN
                nn.init.constant_(m.weight,1)
                nn.init.constant_(m.bias,0)
            elif isinstance(m,nn.Linear): # lnit dense
                nn.init.kaiming_normal_(m.weight)
                nn.init.zeros_(m.bias)
            
    def forward(self,x):
        return self.net(x)

C = ConvolutionalNeuralNetworkClass(
    name='cnn',xdim=[1,28,28],ksize=3,cdims=[32,64],
    hdims=[32],ydim=10).to(device)
loss = nn.CrossEntropyLoss()
optm = optim.Adam(C.parameters(),lr=1e-3)