Mask RCNN
是作者Kaiming He
於2018
年發表的論文
ROI Pooling 和 ROI Align 的區別
Understanding Region of Interest — (RoI Align and RoI Warp)
Mask R-CNN 網路結構
Mask RCNN
繼承自 Faster RCNN
主要有三個改進:
feature map
的提取採用了FPN
的多尺度特徵網路ROI Pooling
改進為ROI Align
- 在
RPN
後面,增加了採用FCN
結構的mask
分割分支
網路結構如下圖所示:
可以看出,Mask RCNN 是一種先檢測物體,再分割的思路,簡單直接,在建模上也更有利於網路的學習。
骨幹網路 FPN
卷積網路的一個重要特徵:深層網路容易響應語義特徵,淺層網路容易響應影像特徵。Mask RCNN
的使用了 ResNet
和 FPN
結合的網路作為特徵提取器。
FPN
的程式碼出現在 ./mrcnn/model.py
中,核心程式碼如下:
if callable(config.BACKBONE):
_, C2, C3, C4, C5 = config.BACKBONE(input_image, stage5=True,
train_bn=config.TRAIN_BN)
else:
_, C2, C3, C4, C5 = resnet_graph(input_image, config.BACKBONE,
stage5=True, train_bn=config.TRAIN_BN)
# Top-down Layers
# TODO: add assert to varify feature map sizes match what's in config
P5 = KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (1, 1), name='fpn_c5p5')(C5)
P4 = KL.Add(name="fpn_p4add")([
KL.UpSampling2D(size=(2, 2), name="fpn_p5upsampled")(P5),
KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (1, 1), name='fpn_c4p4')(C4)])
P3 = KL.Add(name="fpn_p3add")([
KL.UpSampling2D(size=(2, 2), name="fpn_p4upsampled")(P4),
KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (1, 1), name='fpn_c3p3')(C3)])
P2 = KL.Add(name="fpn_p2add")([
KL.UpSampling2D(size=(2, 2), name="fpn_p3upsampled")(P3),
KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (1, 1), name='fpn_c2p2')(C2)])
# Attach 3x3 conv to all P layers to get the final feature maps.
P2 = KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (3, 3), padding="SAME", name="fpn_p2")(P2)
P3 = KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (3, 3), padding="SAME", name="fpn_p3")(P3)
P4 = KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (3, 3), padding="SAME", name="fpn_p4")(P4)
P5 = KL.Conv2D(config.TOP_DOWN_PYRAMID_SIZE, (3, 3), padding="SAME", name="fpn_p5")(P5)
# P6 is used for the 5th anchor scale in RPN. Generated by
# subsampling from P5 with stride of 2.
P6 = KL.MaxPooling2D(pool_size=(1, 1), strides=2, name="fpn_p6")(P5)
# Note that P6 is used in RPN, but not in the classifier heads.
rpn_feature_maps = [P2, P3, P4, P5, P6]
mrcnn_feature_maps = [P2, P3, P4, P5]
其中 resnet_graph
函式定義如下:
def resnet_graph(input_image, architecture, stage5=False, train_bn=True):
"""Build a ResNet graph.
architecture: Can be resnet50 or resnet101
stage5: Boolean. If False, stage5 of the network is not created
train_bn: Boolean. Train or freeze Batch Norm layers
"""
assert architecture in ["resnet50", "resnet101"]
# Stage 1
x = KL.ZeroPadding2D((3, 3))(input_image)
x = KL.Conv2D(64, (7, 7), strides=(2, 2), name='conv1', use_bias=True)(x)
x = BatchNorm(name='bn_conv1')(x, training=train_bn)
x = KL.Activation('relu')(x)
C1 = x = KL.MaxPooling2D((3, 3), strides=(2, 2), padding="same")(x)
# Stage 2
x = conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1), train_bn=train_bn)
x = identity_block(x, 3, [64, 64, 256], stage=2, block='b', train_bn=train_bn)
C2 = x = identity_block(x, 3, [64, 64, 256], stage=2, block='c', train_bn=train_bn)
# Stage 3
x = conv_block(x, 3, [128, 128, 512], stage=3, block='a', train_bn=train_bn)
x = identity_block(x, 3, [128, 128, 512], stage=3, block='b', train_bn=train_bn)
x = identity_block(x, 3, [128, 128, 512], stage=3, block='c', train_bn=train_bn)
C3 = x = identity_block(x, 3, [128, 128, 512], stage=3, block='d', train_bn=train_bn)
# Stage 4
x = conv_block(x, 3, [256, 256, 1024], stage=4, block='a', train_bn=train_bn)
block_count = {"resnet50": 5, "resnet101": 22}[architecture]
for i in range(block_count):
x = identity_block(x, 3, [256, 256, 1024], stage=4, block=chr(98 + i), train_bn=train_bn)
C4 = x
# Stage 5
if stage5:
x = conv_block(x, 3, [512, 512, 2048], stage=5, block='a', train_bn=train_bn)
x = identity_block(x, 3, [512, 512, 2048], stage=5, block='b', train_bn=train_bn)
C5 = x = identity_block(x, 3, [512, 512, 2048], stage=5, block='c', train_bn=train_bn)
else:
C5 = None
return [C1, C2, C3, C4, C5]
anchor 錨框生成規則
在 Faster-RCNN 中可以將 SCALE
也可以設定為多個值,而在 Mask RCNN 中則是每一特徵層只對應著一個SCALE
即對應著上述所設定的 16。
實驗
何凱明在論文中做了很多對比單個模組試驗,並放出了對比結果表格。
從上圖表格可以看出:
sigmoid
和softmax
對比,sigmoid
有不小提升;- 特徵網路選擇:可以看出更深的網路和採用
FPN
的實驗效果更好,可能因為 FPN 綜合考慮了不同尺寸的feature map
的資訊,因此能夠把握一些更精細的細節。 RoI Align
和RoI Pooling
對比:在 instance segmentation 和 object detection 上都有不小的提升。這樣看來,RoIAlign 其實就是一個更加精準的 RoIPooling,把前者放到 Faster RCNN 中,對結果的提升應該也會有幫助。