onnx模型解析
1.原始模型分析
由于centerface的模型是onnx的,可以通过netron工具包查看改模型的输入输出维度
import netron modelPath = ""centerface.onnx"" netron.start(modelPath)
运行上述代码会生成如下的网络结构图:
点击input.1可以查看模型的输入输出维度,如下图
从上图可以看出原始模型的输入维度是(10,3,32,32)输出537的维度是(10,1,8,8);538的维度是(10,2,8,8);539的维度是(10,2,8,8);540的维度是(10,2,8,8)
2.模型修改
由于我们的实际图片大小是1920*1080,所有要对网络的输入输出维度进行修改
import onnx import math input_size =(1080,1920) model = onnx.load_model("centerface.onnx") d = model.graph.input[0].type.tensor_type.shape.dim print(d) rate = (int(math.ceil(input_size[0]/d[2].dim_value)),int(math.ceil(input_size[1]/d[3].dim_value))) print("rare",rate) d[0].dim_value = 1 d[2].dim_value *= rate[0] d[3].dim_value *= rate[1] for output in model.graph.output: d = output.type.tensor_type.shape.dim print(d) d[0].dim_value = 1 d[2].dim_value *= rate[0] d[3].dim_value *= rate[1] onnx.save_model(model,"centerface_1088_1920.onnx" )
再次查看 centerface_1088_1920.onn模型的输入输出维度如下图:
也可以使用代码方式查看网络结构,代码如下
import onnx from onnx import helper import sys,getopt #加载模型 def loadOnnxModel(path): model = onnx.load(path) return model #获取节点和节点的输入输出名列表,一般节点的输入将来自于上一层的输出放在列表前面,参数放在列表后面 def getNodeAndIOname(nodename,model): for i in range(len(model.graph.node)): if model.graph.node[i].name == nodename: Node = model.graph.node[i] input_name = model.graph.node[i].input output_name = model.graph.node[i].output return Node,input_name,output_name #获取对应输入信息 def getInputTensorValueInfo(input_name,model): in_tvi = [] for name in input_name: for params_input in model.graph.input: if params_input.name == name: in_tvi.append(params_input) for inner_output in model.graph.value_info: if inner_output.name == name: in_tvi.append(inner_output) return in_tvi #获取对应输出信息 def getOutputTensorValueInfo(output_name,model): out_tvi = [] for name in output_name: out_tvi = [inner_output for inner_output in model.graph.value_info if inner_output.name == name] if name == model.graph.output[0].name: out_tvi.append(model.graph.output[0]) return out_tvi #获取对应超参数值 def getInitTensorValue(input_name,model): init_t = [] for name in input_name: init_t = [init for init in model.graph.initializer if init.name == name] return init_t #构建单个节点onnx模型 def createSingelOnnxModel(ModelPath,nodename,SaveType="",SavePath=""): model = loadOnnxModel(str(ModelPath)) Node,input_name,output_name = getNodeAndIOname(nodename,model) in_tvi = getInputTensorValueInfo(input_name,model) out_tvi = getOutputTensorValueInfo(output_name,model) init_t = getInitTensorValue(input_name,model) graph_def = helper.make_graph( [Node], nodename, inputs=in_tvi, # 输入 outputs=out_tvi, # 输出 initializer=init_t, # initalizer ) model_def = helper.make_model(graph_def, producer_name='onnx-example') print(nodename+"onnx模型生成成功!")
#获取节点数量 def getNodeNum(model): return len(model.graph.node) #获取节点类型 def getNodetype(model): op_name = [] for i in range(len(model.graph.node)): if model.graph.node[i].op_type not in op_name: op_name.append(model.graph.node[i].op_type) return op_name #获取节点名列表 def getNodeNameList(model): NodeNameList = [] for i in range(len(model.graph.node)): NodeNameList.append(model.graph.node[i].name) return NodeNameList #获取模型的输入信息 def getModelInputInfo(model): return model.graph.input[0] #获取模型的输出信息 def getModelOutputInfo(model): return model.graph.output[0:4]