在caffe中layer分爲以下幾類:
- Data Layers
- Vision Layers
- Recurrent Layers
- Common Layers
- Normalization Layers
- Activation / Neuron Layers
- Utility Layers
Loss Layers
首先介紹數據層,數據層是每個模型的最底層,是模型的入口,不僅提供數據的輸入,也提供數據從Blobs轉換成別的格式進行保存輸出。通常數據的預處理(如減去均值, 放大縮小, 裁剪和鏡像等),也在這一層設置參數實現。數據來源可以來自高效的數據庫(如LevelDB和LMDB),也可以直接來自於內存。如果不是很注重效率的話,數據也可來自磁盤的hdf5文件和圖片格式文件。所有的數據層的都具有的公用參數:先看示例:
layer {
name: "cifar"
type: "Data"
top: "data"
top: "label"
include {
phase: TRAIN
}
transform_param {
mean_file: "examples/cifar10/mean.binaryproto"
}
data_param {
source: "examples/cifar10/cifar10_train_lmdb"
batch_size: 100
backend: LMDB
}
}
下面分別介紹這些參數的含義:
1)name:表示該層的名字,一般選取有意義的名稱;
2) type: 表示層類型,如果是Data,表示數據來源於LevelDB或LMDB。根據數據的來源不同,數據層的類型也不同(後面會詳細闡述)。通常都是採 用的LevelDB或LMDB數據,因此層類型設置爲Data。
3)top或bottom: 每一層用bottom來輸入數據,用top來輸出數據。如果只有top沒有bottom,則此層只有輸出,沒有輸入。反之亦然。如果有多個 top或多個bottom,表示有多個blobs數據的輸入和輸出。
4)data 與 label: 在數據層中,至少有一個命名爲data的top。如果有第二個top,若是分類問題則一般是label。根據任務的不同top的有所不同,例如:siftflow-fcn16s 的datalayer如下:
layer {
name: "data"
type: "Python"
top: "data"
top: "sem"
top: "geo"
python_param {
module: "siftflow_layers"
layer: "SIFTFlowSegDataLayer"
param_str: "{\'siftflow_dir\': \'../data/sift-flow\', \'seed\': 1337, \'split\': \'trainval\'}"
}
}
5)include: 一般訓練的時候和測試的時候,模型的層是不一樣的。該層(layer)是屬於訓練階段的層,還是屬於測試階段的層,需要用include來指定。如果沒有include參數,則表示該層既在訓練模型中,又在測試模型中
6)Transform_param: 數據的預處理,可以將數據變換到定義的範圍內。如設置scale、mirror、crop_size、mean_file等。例如:
transform_param {
scale: 0.00390625
mean_file_size: "examples/cifar10/mean.binaryproto"
# 用一個配置文件來進行均值操作
mirror: 1 # 1表示開啓鏡像,0表示關閉,也可用ture和false來表示
# 剪裁一個 227*227的圖塊,在訓練階段隨機剪裁,在測試階段從中間裁剪
crop_size: 227
}
參考caffe.proto 文件,有關transform_param 的code如下:
// Message that stores parameters used to apply transformation
// to the data layer's data
message TransformationParameter {
// For data pre-processing, we can do simple scaling and subtracting the
// data mean, if provided. Note that the mean subtraction is always carried
// out before scaling.
optional float scale = 1 [default = 1];
// Specify if we want to randomly mirror data.
optional bool mirror = 2 [default = false];
// Specify if we would like to randomly crop an image.
optional uint32 crop_size = 3 [default = 0];
// mean_file and mean_value cannot be specified at the same time
optional string mean_file = 4;
// if specified can be repeated once (would subtract it from all the channels)
// or can be repeated the same number of times as channels
// (would subtract them from the corresponding channel)
repeated float mean_value = 5;
// Force the decoded image to have 3 color channels.
optional bool force_color = 6 [default = false];
// Force the decoded image to have 1 color channels.
optional bool force_gray = 7 [default = false];
}
type:參數輸入類型很多,所以單獨拿出來做解釋,下面詳細介紹:
1) tpye:Data,這種最常用 ,例如:
layer {
name: "mnist"
type: "Data"
top: "data"
top: "label"
include {
phase: TRAIN
}
transform_param {
scale: 0.00390625
}
data_param {
source: "examples/mnist/mnist_train_lmdb"
batch_size: 64
backend: LMDB
}
}
必須設置的參數:
source: 包含數據庫的目錄名稱,如examples/mnist/mnist_train_lmdb
batch_size: 每次處理的數據個數,如64
可選的參數:
rand_skip: 在開始的時候,路過某個數據的輸入。通常對異步的SGD很有用。
backend: 選擇是採用LevelDB還是LMDB, 默認是LevelDB.
參考源碼:
**message DataParameter {
enum DB {
LEVELDB = 0;
LMDB = 1;
}
// Specify the data source.
optional string source = 1;
// Specify the batch size.
optional uint32 batch_size = 4;
// The rand_skip variable is for the data layer to skip a few data points
// to avoid all asynchronous sgd clients to start at the same point. The skip
// point would be set as rand_skip * rand(0,1). Note that rand_skip should not
// be larger than the number of keys in the database.
// DEPRECATED. Each solver accesses a different subset of the database.
optional uint32 rand_skip = 7 [default = 0];
optional DB backend = 8 [default = LEVELDB];
// DEPRECATED. See TransformationParameter. For data pre-processing, we can do
// simple scaling and subtracting the data mean, if provided. Note that the
// mean subtraction is always carried out before scaling.
optional float scale = 2 [default = 1];
optional string mean_file = 3;
// DEPRECATED. See TransformationParameter. Specify if we would like to randomly
// crop an image.
optional uint32 crop_size = 5 [default = 0];
// DEPRECATED. See TransformationParameter. Specify if we want to randomly mirror
// data.
optional bool mirror = 6 [default = false];
// Force the encoded image to have 3 color channels
optional bool force_encoded_color = 9 [default = false];
// Prefetch queue (Increase if data feeding bandwidth varies, within the
// limit of device memory for GPU training)
optional uint32 prefetch = 10 [default = 4];
}**
2)type:memory_data
layer {
top: "data"
top: "label"
name: "memory_data"
type: "MemoryData"
memory_data_param{
batch_size: 2
height: 100
width: 100
channels: 1
}
transform_param {
scale: 0.0078125
mean_file: "mean.proto"
mirror: false
}
}
必須設置的參數: batch_size:每一次處理的數據個數,比如2; channels:通道數; height:高度; width: 寬度;
參考源碼:
message MemoryDataParameter {
optional uint32 batch_size = 1;
optional uint32 channels = 2;
optional uint32 height = 3;
optional uint32 width = 4;
}
3) type:HDF5data
layer {
name: "data"
type: "HDF5Data"
top: "data"
top: "label"
hdf5_data_param {
source: "examples/hdf5_classification/data/train.txt"
batch_size: 10
}
}
必須設置的參數:source: 讀取的文件名稱;batch_size: 每一次處理的數據個數
// Message that stores parameters used by HDF5DataLayer
message HDF5DataParameter {
// Specify the data source.
optional string source = 1;
// Specify the batch size.
optional uint32 batch_size = 2;
// Specify whether to shuffle the data.
// If shuffle == true, the ordering of the HDF5 files is shuffled,
// and the ordering of data within any given HDF5 file is shuffled,
// but data between different files are not interleaved; all of a file's
// data are output (in a random order) before moving onto another file.
optional bool shuffle = 3 [default = false];
}
4) type:ImageData
layer {
name: "data"
type: "ImageData"
top: "data"
top: "label"
transform_param {
mirror: false
crop_size: 227
mean_file: "data/ilsvrc12/imagenet_mean.binaryproto"
}
image_data_param {
source: "examples/_temp/file_list.txt"
batch_size: 50
new_height: 256
new_width: 256
}
}
必須設置的參數:
source: 一個文本文件的名字,每一行給定一個圖片文件的名稱和標籤(label)
batch_size: 每一次處理的數據個數,即圖片數
可選參數:
rand_skip: 在開始的時候,路過某個數據的輸入。通常對異步的SGD很有用。
shuffle: 隨機打亂順序,默認值爲false
new_height,new_width: 如果設置,則將圖片進行resize
參考源碼:
message ImageDataParameter {
// Specify the data source.
optional string source = 1;
// Specify the batch size.
optional uint32 batch_size = 4 [default = 1];
// The rand_skip variable is for the data layer to skip a few data points
// to avoid all asynchronous sgd clients to start at the same point. The skip
// point would be set as rand_skip * rand(0,1). Note that rand_skip should not
// be larger than the number of keys in the database.
optional uint32 rand_skip = 7 [default = 0];
// Whether or not ImageLayer should shuffle the list of files at every epoch.
optional bool shuffle = 8 [default = false];
// It will also resize images if new_height or new_width are not zero.
optional uint32 new_height = 9 [default = 0];
optional uint32 new_width = 10 [default = 0];
// Specify if the images are color or gray
optional bool is_color = 11 [default = true];
// DEPRECATED. See TransformationParameter. For data pre-processing, we can do
// simple scaling and subtracting the data mean, if provided. Note that the
// mean subtraction is always carried out before scaling.
optional float scale = 2 [default = 1];
optional string mean_file = 3;
// DEPRECATED. See TransformationParameter. Specify if we would like to randomly
// crop an image.
optional uint32 crop_size = 5 [default = 0];
// DEPRECATED. See TransformationParameter. Specify if we want to randomly mirror
// data.
optional bool mirror = 6 [default = false];
optional string root_folder = 12 [default = ""];
}
5)type:Python 這個有點牛,可以自己定義
例如:鼎鼎大名的fcn 就採用了這個:
layer {
name: "data"
type: "Python"
top: "data"
top: "sem"
top: "geo"
python_param {
module: "siftflow_layers"
layer: "SIFTFlowSegDataLayer"
param_str: "{\'siftflow_dir\': \'../data/sift-flow\', \'seed\': 1337, \'split\': \'trainval\'}"
}
}
具體可去caffe官網查看內容,先寫這麼多吧!!!