本我的项目利用MindSpore框架搭建YOLOv3指标检测模型，从PASCAL VOC 2012数据集中提取出的人体指标检测数据进行模型训练，失去一个人体指标检测模型。冀望通过本次我的项目为MindSpore生态尽本人的一份绵薄之力。1.环境筹备抉择MindSpore版本为1.5或1.6，硬件为GPU。能够参照 https://www.mindspore.cn/install 依据本人的本地环境进行装置。笔者应用了华为云->ModelArt->开发环境->notebook，用这个产品的益处是它的环境包含MindSpore框架都曾经装好了，笔者抉择的规格是 GPU: 1*V100(32GB)|CPU: 8核 64GB ，这种规格大略要花100~200元（具体我忘了，兴许更少）能跑完本我的项目，而且速度还挺快。降级MindSpore版本同样可参考 https://www.mindspore.cn/install 。2.数据集解决PASCAL VOC 2012数据集蕴含训练集 5717 张，验证集 5823 张，共有20个检测的类别。2.1 提取出"person"指标检测数据，对检测框作聚类咱们首先要从数据集中提取带"person"指标的图片，而后遍历"person"指标检测框，以检测框长和宽为坐标选出9个聚类中心点，这9个中心点的坐标将用于作为YOLOv3的先验框大小。这部分内容的代码能够参考 附件\choose_person 文件夹，将这个文件夹下的python代码放到voc2012数据集的 VOCtrainval_11-May-2012\VOCdevkit 目录（该目录中有一个VOC2012文件夹，外面有ImageSets等目录），挨个运行cluster.py、rand_choose.py即可（可参考附件中的readme.txt）。在执行完两个python程序后，会失去相似于voc2012数据集 VOCtrainval_11-May-2012\VOCdevkit\VOC2012\ImageSets\Main\ 目录下的 train.txt、val.txt 的 txt 文本文件，将它们搁置于上述的 Main\ 目录下。2.2 数据集加载MindSpore 为咱们提供了加载 PASCAL VOC 数据集的接口 VOCDataset，可参考 MindSpore 官网中的文档应用。这里咱们再为它提供一些配套的解决和封装（参见 附件\dataset\voc2012_dataset.py ）："""读取VOC2012数据集"""
import mindspore.dataset as ds
import mindspore.dataset.vision.c_transforms as CV
import cv2
import numpy as np
附件\dataset\transforms.py
from transforms import reshape_fn, MultiScaleTrans
def create_voc2012_dataset(config, cv_num):
"""create VOC2012 dataset"""voc2012_dat = ds.VOCDataset(dataset_dir=config.data_path, task="Detection", usage=config.data_usage, shuffle=config.data_training, num_parallel_workers=8)dataset_size = voc2012_dat.get_dataset_size()config.class_to_idx = voc2012_dat.get_class_indexing()cv2.setNumThreads(0)if config.data_training: multi_scale_trans = MultiScaleTrans(config, cv_num) dataset_input_column_names = ["image", "bbox", "label", "truncate", "difficult"] dataset_output_column_names = ["image", "annotation", "bbox1", "bbox2", "bbox3", "gt_box1", "gt_box2", "gt_box3"] voc2012_dat = voc2012_dat.map(operations=CV.Decode(), input_columns=["image"]) voc2012_dat = voc2012_dat.batch(config.batch_size, per_batch_map=multi_scale_trans, input_columns=dataset_input_column_names, output_columns=dataset_output_column_names, num_parallel_workers=8, drop_remainder=True) voc2012_dat = voc2012_dat.repeat(config.max_epoch-config.pretrained_epoch_num) else: img_id = np.array(range(0,dataset_size)) img_id = img_id.reshape((-1,1)) img_id = ds.GeneratorDataset(img_id, ['img_id'], shuffle=False) voc2012_dat = voc2012_dat.zip(img_id) compose_map_func = (lambda image, img_id: reshape_fn(image, img_id, config)) voc2012_dat = voc2012_dat.map(operations=CV.Decode(), input_columns=["image"], num_parallel_workers=8) voc2012_dat = voc2012_dat.map(operations=compose_map_func, input_columns=["image", "img_id"], output_columns=["image", "image_shape", "img_id"], column_order=["image", "image_shape", "img_id"], num_parallel_workers=8) hwc_to_chw = CV.HWC2CHW() voc2012_dat = voc2012_dat.map(operations=hwc_to_chw, input_columns=["image"], num_parallel_workers=8) voc2012_dat = voc2012_dat.batch(config.batch_size, drop_remainder=True) voc2012_dat = voc2012_dat.repeat(1) return voc2012_dat, dataset_size这个函数大略的意思是将 VOCDataset 读出的数据集解决成咱们想要的样子，即最初返回的 voc2012_dat。dataset_size 是读出的数据集大小（图片张数）。对于训练来说（config.data_training==True），首先要将数据集解决成batch，并能以 MindSpore 的 Tensor 类型返回，这要求将一个 batch 中的图片reshape成对立大小，其中每张输出图像都对应三种尺度的输入特色图，即上方函数中的"bbox1"、"bbox2"、"bbox3"，函数 create_voc2012_dataset 要做的就是将数据集中的标签检测框数据依据坐标和大小映射到对应尺度的特色图中（办法是将标签检测框与选出的先验框进行IoU计算，标签检测框落入IoU得分大的对应先验框地位，详见 附件\dataset\transforms.py 中的 _preprocess_true_boxes 函数），最初 "gt_box" 以Tensor模式寄存所有标签检测框，不便后续的计算。用于训练的图片还须要进行图像增强操作，包含对图像随机缩放、翻转、旋转、剪切、平移以及色彩随机变换，相应的标签检测框地位也要变换，这些操作都在 附件\dataset\transforms.py 文件（该文件来自于 https://gitee.com/mindspore/models/blob/r1.5/official/cv/yolov3_darknet53/src/transforms.py ，我只改了大量内容，例如增加代码将VOCDataset读出的标签检测框格局[x y w h]解决成了[xmin ymin xmax ymax]）中，并最终在 MultiScaleTrans 中调用。对于测试，同样能够将图片解决成batch，也要reshape成对立大小，另外还要保留原图像的 shape （即上方函数的"image_shape"），不便将测试时模型推理失去的坐标映射到原图中。"img_id" 是从0到n-1（设数据集大小，即dataset_size为n）的数字，代表图片对应于 config.data_usage 所指向的文件（这个文件即2.1中生成的 txt 文件）的第几行，从而进一步失去对应是哪张图片。参数 config 的具体含意和设置，以及数据集的具体应用办法，请见前面的模型训练和模型测试局部。3.模型搭建本我的项目采纳Darknet53作为YOLOv3的骨干网络，Darknet53和YOLOv3模型的结构图如下：
上面咱们开始构建模型，本我的项目参考了 https://gitee.com/mindspore/m... ，寻着目录往上找还能找到许多MindSpore写的常见模型，个别大家将数据处理成他们的格局间接用他们的模型就行了。3.1 Darknet53"""YOLOv3 backbone: darknet53"""
import mindspore.nn as nn
from mindspore.ops import operations as P
def conv_block(in_channels,
out_channels, kernel_size, stride, dilation=1):"""Get a conv2d batchnorm and relu layer"""pad_mode = 'same'padding = 0return nn.SequentialCell( [nn.Conv2d(in_channels, out_channels, kernel_size=kernel_size, stride=stride, padding=padding, dilation=dilation, pad_mode=pad_mode), nn.BatchNorm2d(out_channels, momentum=0.1), nn.ReLU()])class ResidualBlock(nn.Cell):
"""DarkNet V1 residual block definition.Args: in_channels: Integer. Input channel. out_channels: Integer. Output channel.Returns: Tensor, output tensor.Examples: ResidualBlock(3, 208)"""expansion = 4def __init__(self, in_channels, out_channels): super(ResidualBlock, self).__init__() out_chls = out_channels//2 self.conv1 = conv_block(in_channels, out_chls, kernel_size=1, stride=1) self.conv2 = conv_block(out_chls, out_channels, kernel_size=3, stride=1) self.add = P.Add()def construct(self, x): identity = x out = self.conv1(x) out = self.conv2(out) out = self.add(out, identity) return outclass DarkNet(nn.Cell):
"""DarkNet V1 network.Args: block: Cell. Block for network. layer_nums: List. Numbers of different layers. in_channels: Integer. Input channel. out_channels: Integer. Output channel. detect: Bool. Whether detect or not. Default:False.Returns: Tuple, tuple of output tensor,(f1,f2,f3,f4,f5).Examples: DarkNet(ResidualBlock, [1, 2, 8, 8, 4], [32, 64, 128, 256, 512], [64, 128, 256, 512, 1024], 100)"""def __init__(self, block, layer_nums, in_channels, out_channels, detect=False): super(DarkNet, self).__init__() self.outchannel = out_channels[-1] self.detect = detect if not len(layer_nums) == len(in_channels) == len(out_channels) == 5: raise ValueError("the length of layer_num, inchannel, outchannel list must be 5!") self.conv0 = conv_block(3, in_channels[0], kernel_size=3, stride=1) self.conv1 = conv_block(in_channels[0], out_channels[0], kernel_size=3, stride=2) self.layer1 = self._make_layer(block, layer_nums[0], in_channel=out_channels[0], out_channel=out_channels[0]) self.conv2 = conv_block(in_channels[1], out_channels[1], kernel_size=3, stride=2) self.layer2 = self._make_layer(block, layer_nums[1], in_channel=out_channels[1], out_channel=out_channels[1]) self.conv3 = conv_block(in_channels[2], out_channels[2], kernel_size=3, stride=2) self.layer3 = self._make_layer(block, layer_nums[2], in_channel=out_channels[2], out_channel=out_channels[2]) self.conv4 = conv_block(in_channels[3], out_channels[3], kernel_size=3, stride=2) self.layer4 = self._make_layer(block, layer_nums[3], in_channel=out_channels[3], out_channel=out_channels[3]) self.conv5 = conv_block(in_channels[4], out_channels[4], kernel_size=3, stride=2) self.layer5 = self._make_layer(block, layer_nums[4], in_channel=out_channels[4], out_channel=out_channels[4])def _make_layer(self, block, layer_num, in_channel, out_channel): """ Make Layer for DarkNet. :param block: Cell. DarkNet block. :param layer_num: Integer. Layer number. :param in_channel: Integer. Input channel. :param out_channel: Integer. Output channel. Examples: _make_layer(ConvBlock, 1, 128, 256) """ layers = [] darkblk = block(in_channel, out_channel) layers.append(darkblk) for _ in range(1, layer_num): darkblk = block(out_channel, out_channel) layers.append(darkblk) return nn.SequentialCell(layers)def construct(self, x): c1 = self.conv0(x) c2 = self.conv1(c1) c3 = self.layer1(c2) c4 = self.conv2(c3) c5 = self.layer2(c4) c6 = self.conv3(c5) c7 = self.layer3(c6) c8 = self.conv4(c7) c9 = self.layer4(c8) c10 = self.conv5(c9) c11 = self.layer5(c10) if self.detect: return c7, c9, c11 return c11def get_out_channels(self): return self.outchanneldef get_darknet53(detect=False):
"""Get DarkNet53 neural network.Returns: Cell, cell instance of DarkNet53 neural network.Examples: darknet53()"""return DarkNet(ResidualBlock, [1, 2, 8, 8, 4], [32, 64, 128, 256, 512], [64, 128, 256, 512, 1024], detect)（未完，请见 YOLOv3人体指标检测模型实现（二））