PP-PicoDet是轻量级实时移动端目标检测模型，我们提出了从小到大的一系列模型，包括S、M、L等，超越现有SOTA模型。

模型特色：

• 🌟精度高：1M参数量以内mAP(0.5:0.95)达到30.6，3.3M参数量mAP(0.5:0.95)达到40.9。
• 🚀速度快：在SD865上达到150FPS。
• 😊部署友好：我们支持PaddleInference/PaddleLite/MNN/NCNN/OpenVINO，并且提供C++/Python/Android demo。

链接：

• 详细算法细节请参考paper：https://arxiv.org/abs/2111.00902
• Readme&配置文件： https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/picodet

欢迎大家试用，有疑问欢迎讨论盖楼~

和其他模型对比：

FAQ汇总： （持续更新中）

• 版本要求： 训练导出模型要求Paddle版本统一，同时 PaddlePaddle >= 2.1.2。
• 学习率、GPU数和batch-size关系： 采用线性伸缩准则，发布的配置文件基本都是4卡GPU训练的，例如：变成单卡，请学习率除以4，如果batch size从80变成40，请学习率再除以2。
• 配置优先级： 一般picodet_x_coco.yml中的配置优先级高于__base__中配置，picodet_x_coco.yml中的所有设置会覆盖__base__中配置，所以修改picodet_x_coco.yml的配置即可。
• 在自己数据集上训练模型： 支持COCO和VOC两种数据格式，同时建议采用迁移学习加快收敛，具体步骤：从PicoDet的Readme中拷贝COCO上训好的pretrain weights链接，更新配置文件中pretrain_weights参数为COCO上训好的权重。

为了方便大家交流沟通，欢迎扫码添加微信群，继续交流有关PP-PicoDet的使用及建议~

https://paddledetection.readthedocs.io/advanced_tutorials/inference/docs/windows_vs2019_build.html 我根据上面的说明进行部署，发现报错。 CMakeCache.txt 指向 D:\1.6.1\paddle。 此目录不存在。 后来我打开下面网址https://paddledetection.readthedocs.io/advanced_tutorials/inference/docs/windows_vs2015_build.html 这个部署版本与paddleDetection 不同，还是需要1.6版本，我是应该部署1.6版本吗，可我的训练文件是0.2的paddleDetection。请问我应该如何解决。

问题确认 Search before asking

• [X] 我已经查询历史issue，没有报过同样bug。I have searched the issues and found no similar bug report.

bug描述 Describe the Bug

@yghstill TinyPose模型的自动化压缩代码整理完成了：https://github.com/PaddlePaddle/PaddleSlim/tree/develop/demo/auto_compression/detection 配置文件是：configs/tinypose_qat_dis.yaml，启动方式和readme中完全一致，如果全量化的话可以先使用这个 上面是你说的话，可是出现了几个问题。 1 你给出的这个链接 404打不开了，于是我找到了其他的位置，找到你说的yml，运行程序 2 python3 -m paddle.distributed.launch --log_dir=log0705 --gpus 0,1 run.py --config_path=./configs/tinypose_qat_dis.yaml --save_dir='./output0705/'
python3 run.py --config_path=./configs/tinypose_qat_dis.yaml --save_dir='./output0705/'
以上两种方式都尝试了，可是报错如下 请大佬能认真帮忙看下吗？ 3 是否有真正的可复现而且无bug的版本发出来，是否可以进行自测再发出来，多谢

BR

复现环境 Environment

       paddle-gpu 2.23 以上
       paddledet release/2.4
       paddleslim develop
       cuda 11.3
      pytorch 跟cuda配套

是否愿意提交PR Are you willing to submit a PR?

• [x] Yes I'd like to help by submitting a PR!

训练yoloe_s报错,提示：

ValueError: (InvalidArgument) Broadcast dimension mismatch. Operands could not be broadcast together with the shape of 
X = [1, 3024, 2] and the shape of Y = [8400, 2]. Received [3024] in X is not equal to [8400] in Y at i:1.
  [Hint: Expected x_dims_array[i] == y_dims_array[i] || x_dims_array[i] <= 1 || y_dims_array[i] <= 1 == true, 
but received x_dims_array[i] == y_dims_array[i] || x_dims_array[i] <= 1 || y_dims_array[i] <= 1:0 != true:1.] 
(at /paddle/paddle/fluid/operators/elementwise/elementwise_op_function.h:240)
  [operator < elementwise_add > error]

配置文件在 mask_rcnn_r50_2x.yml 的基础上进行修改，执行下面的命令之后：

!python tools/train.py -c configs/myconfig/mask_rcnn_r50_2x.yml --eval -o use_gpu=true --use_vdl=True --vdl_log_dir=vdl_dir/scalar

报错：

Traceback (most recent call last):
  File "tools/train.py", line 377, in <module>
    main()
  File "tools/train.py", line 146, in main
    fetches = model.eval(feed_vars)
  File "/home/aistudio/work/PaddleDetection/ppdet/modeling/architectures/mask_rcnn.py", line 338, in eval
    return self.build(feed_vars, 'test')
  File "/home/aistudio/work/PaddleDetection/ppdet/modeling/architectures/mask_rcnn.py", line 81, in build
    self._input_check(required_fields, feed_vars)
  File "/home/aistudio/work/PaddleDetection/ppdet/modeling/architectures/mask_rcnn.py", line 271, in _input_check
    "{} has no {} field".format(feed_vars, var)
AssertionError: OrderedDict([('image', name: "image"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: FP32
      dims: -1
      dims: 3
      dims: -1
      dims: -1
    }
    lod_level: 0
  }
}
persistable: false
need_check_feed: true
), ('im_info', name: "im_info"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: FP32
      dims: -1
      dims: 3
    }
    lod_level: 0
  }
}
persistable: false
need_check_feed: true
), ('im_id', name: "im_id"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: INT64
      dims: -1
      dims: 1
    }
    lod_level: 0
  }
}
persistable: false
need_check_feed: true
), ('gt_bbox', name: "gt_bbox"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: FP32
      dims: -1
      dims: 4
    }
    lod_level: 1
  }
}
persistable: false
need_check_feed: true
), ('gt_class', name: "gt_class"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: INT32
      dims: -1
      dims: 1
    }
    lod_level: 1
  }
}
persistable: false
need_check_feed: true
), ('is_crowd', name: "is_crowd"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: INT32
      dims: -1
      dims: 1
    }
    lod_level: 1
  }
}
persistable: false
need_check_feed: true
), ('gt_mask', name: "gt_mask"
type {
  type: LOD_TENSOR
  lod_tensor {
    tensor {
      data_type: FP32
      dims: -1
      dims: 2
    }
    lod_level: 3
  }
}
persistable: false
need_check_feed: true
)]) has no im_shape field

我查看了 mask_rcnn.py 第81 行，这里应该是因为 model != 'train' 所以 self._input_check(required_fields, feed_vars) check 了 im_shape 我的训练启动命令有问题吗？

问题确认 Search before asking

• [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.4/configs/dota 请问这个链接的镜像是不是没了。 我显卡3070 cuda11.1 paddle2.2 paddledetection 2.2 按照教程训练显示cudnn错误。会不会是目前这个只支持cuda10 cudnn7.可30显卡好像对cuda10不太友好。请问有没有办法能本地安装上这个模型阿

ERROR - DataLoader reader thread raised an exception!                                 
Exception in thread Thread-1:                                                                                   
Traceback (most recent call last):                                                                              
  File "/usr/local/python3/lib/python3.6/threading.py", line 916, in _bootstrap_inner                           
    self.run()                                                                                                  
  File "/usr/local/python3/lib/python3.6/threading.py", line 864, in run                                        
    self._target(*self._args, **self._kwargs)                                                                   
  File "/home/.local/lib/python3.6/site-packages/paddle/fluid/dataloader/dataloader_iter.py", line 662, 
in _thread_loop                                                                                                 
    six.reraise(*sys.exc_info())                                                                                
  File "/home/.local/lib/python3.6/site-packages/six.py", line 719, in reraise                          
    raise value                                                                                                 
  File "/home/.local/lib/python3.6/site-packages/paddle/fluid/dataloader/dataloader_iter.py", line 650, 
in _thread_loop                                                                                                 
    tmp.set(slot, core.CPUPlace())                                                                              
ValueError: (InvalidArgument) Input object type error or incompatible array data type. tensor.set() supports arr
ay with bool, float16, float32, float64, int8, int16, int32, int64, uint8 or uint16, please check your input or 
input array data type. (at /paddle/paddle/fluid/pybind/tensor_py.h:355)  

这里应该填入的是什么数据啊

问题确认 Search before asking

• [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

ppyoloe_crn_s_300e_coco VOC 数据集

python tools/train.py -c configs/ppyoloe/ppyoloe_crn_s_300e_coco.yml

W0823 14:30:26.446256  4452 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 8.6, Driver API Version: 11.6, Runtime API Version: 11.2
W0823 14:30:26.461884  4452 gpu_resources.cc:91] device: 0, cuDNN Version: 8.2.
[08/23 14:30:27] ppdet.utils.checkpoint INFO: Finish loading model weights: C:\Users\fujunnnn/.cache/paddle/weights\CSPResNetb_s_pretrained.pdparams
[08/23 14:30:30] ppdet.engine INFO: Epoch: [0] [  0/339] learning_rate: 0.000000 loss: 1931307253760.000000 loss_cls: 0.594841 loss_iou: 772522901504.000000 loss_dfl: 5885.125977 loss_l1: 0.105123 eta: 4 days, 9:30:32 batch_cost: 3.7348 data_cost: 0.2500 ips: 2.6775 images/s
Traceback (most recent call last):
  File "tools/train.py", line 177, in <module>
    main()
  File "tools/train.py", line 173, in main
    run(FLAGS, cfg)
  File "tools/train.py", line 127, in run
    trainer.train(FLAGS.eval)
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\engine\trainer.py", line 454, in train
    outputs = model(data)
  File "E:\anaconda3\envs\PaddleDetection\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in __call__
    return self._dygraph_call_func(*inputs, **kwargs)
  File "E:\anaconda3\envs\PaddleDetection\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func
    outputs = self.forward(*inputs, **kwargs)
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\architectures\meta_arch.py", line 59, in forward
    out = self.get_loss()
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\architectures\yolo.py", line 125, in get_loss
    return self._forward()
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\architectures\yolo.py", line 88, in _forward
    yolo_losses = self.yolo_head(neck_feats, self.inputs)
  File "E:\anaconda3\envs\PaddleDetection\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in __call__
    return self._dygraph_call_func(*inputs, **kwargs)
  File "E:\anaconda3\envs\PaddleDetection\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func
    outputs = self.forward(*inputs, **kwargs)
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 217, in forward
    return self.forward_train(feats, targets)
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 160, in forward_train
    ], targets)
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 355, in get_loss
    assigned_scores_sum)
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 291, in _bbox_loss
    assigned_ltrb_pos) * bbox_weight
  File "E:\PaddleX_GUI_2.1.0_win10\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 256, in _df_loss
    pred_dist, target_left, reduction='none') * weight_left
  File "E:\anaconda3\envs\PaddleDetection\lib\site-packages\paddle\nn\functional\loss.py", line 1723, in cross_entropy
    label_max.item()))
ValueError: Target 25479 is out of upper bound.

python tools/train.py -c configs/ppyoloe/ppyoloe_plus_crn_s_80e_coco.yml

W0823 21:31:38.730271 10200 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 8.6, Driver API Version: 11.6, Runtime API Version: 11.2
W0823 21:31:38.750262 10200 gpu_resources.cc:91] device: 0, cuDNN Version: 8.2.
[08/23 21:31:40] ppdet.utils.checkpoint INFO: The shape [365] in pretrained weight yolo_head.pred_cls.0.bias is unmatched with the shape [4] in model yolo_head.pred_cls.0.bias. And the weight yolo_head.pred_cls.0.bias will not be loaded
[08/23 21:31:40] ppdet.utils.checkpoint INFO: The shape [365, 384, 3, 3] in pretrained weight yolo_head.pred_cls.0.weight is unmatched with the shape [4, 384, 3, 3] in model yolo_head.pred_cls.0.weight. And the weight yolo_head.pred_cls.0.weight will not be loaded
[08/23 21:31:40] ppdet.utils.checkpoint INFO: The shape [365] in pretrained weight yolo_head.pred_cls.1.bias is unmatched with the shape [4] in model yolo_head.pred_cls.1.bias. And the weight yolo_head.pred_cls.1.bias will not be loaded
[08/23 21:31:40] ppdet.utils.checkpoint INFO: The shape [365, 192, 3, 3] in pretrained weight yolo_head.pred_cls.1.weight is unmatched with the shape [4, 192, 3, 3] in model yolo_head.pred_cls.1.weight. And the weight yolo_head.pred_cls.1.weight will not be loaded
[08/23 21:31:40] ppdet.utils.checkpoint INFO: The shape [365] in pretrained weight yolo_head.pred_cls.2.bias is unmatched with the shape [4] in model yolo_head.pred_cls.2.bias. And the weight yolo_head.pred_cls.2.bias will not be loaded
[08/23 21:31:40] ppdet.utils.checkpoint INFO: The shape [365, 96, 3, 3] in pretrained weight yolo_head.pred_cls.2.weight is unmatched with the shape [4, 96, 3, 3] in model yolo_head.pred_cls.2.weight. And the weight yolo_head.pred_cls.2.weight will not be loaded
[08/23 21:31:40] ppdet.utils.checkpoint INFO: Finish loading model weights: C:\Users\MM/.cache/paddle/weights\ppyoloe_crn_s_obj365_pretrained.pdparams
Traceback (most recent call last):
  File "tools/train.py", line 172, in <module>
    main()
  File "tools/train.py", line 168, in main
    run(FLAGS, cfg)
  File "tools/train.py", line 132, in run
    trainer.train(FLAGS.eval)
  File "D:\0SDXX\PaddleDetection\ppdet\engine\trainer.py", line 504, in train
    outputs = model(data)
  File "D:\Anaconda3\envs\PaddleSeg\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in __call__
    return self._dygraph_call_func(*inputs, **kwargs)
  File "D:\Anaconda3\envs\PaddleSeg\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func
    outputs = self.forward(*inputs, **kwargs)
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\architectures\meta_arch.py", line 59, in forward
    out = self.get_loss()
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\architectures\yolo.py", line 124, in get_loss
    return self._forward()
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\architectures\yolo.py", line 88, in _forward
    yolo_losses = self.yolo_head(neck_feats, self.inputs)
  File "D:\Anaconda3\envs\PaddleSeg\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in __call__
    return self._dygraph_call_func(*inputs, **kwargs)
  File "D:\Anaconda3\envs\PaddleSeg\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func
    outputs = self.forward(*inputs, **kwargs)
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 216, in forward
    return self.forward_train(feats, targets)
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 161, in forward_train
    ], targets)
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 354, in get_loss
    assigned_scores_sum)
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 290, in _bbox_loss
    assigned_ltrb_pos) * bbox_weight
  File "D:\0SDXX\PaddleDetection\ppdet\modeling\heads\ppyoloe_head.py", line 255, in _df_loss
    pred_dist, target_left, reduction='none') * weight_left
  File "D:\Anaconda3\envs\PaddleSeg\lib\site-packages\paddle\nn\functional\loss.py", line 1723, in cross_entropy
    label_max.item()))
ValueError: Target 28 is out of upper bound.

根据教程，输入 python tools/train.py -c configs/yolov3/yolov3_mobilenet_v1_roadsign.yml ，只训练一轮就结束了，没有任何报错，环境python3.7，paddledetection是clone的2.1-gpu版本，epoch是默认的12。

我在PaddleDetection/ppdet/modeling/necks/yolo_fpn.py 做了如下修改： 【在forward中加入 block = ChannelAttention(block)； block = SpatialAttention()】

class PPYOLOFPN(nn.Layer):
    __shared__ = ['norm_type', 'data_format']

    def __init__(self,
                 in_channels=[512, 1024, 2048],
                 norm_type='bn',
                 data_format='NCHW',
                 coord_conv=False,
                 conv_block_num=2,
                 drop_block=False,
                 block_size=3,
                 keep_prob=0.9,
                 spp=False):
        """
        PPYOLOFPN layer

        Args:
            in_channels (list): input channels for fpn
            norm_type (str): batch norm type, default bn
            data_format (str): data format, NCHW or NHWC
            coord_conv (bool): whether use CoordConv or not
            conv_block_num (int): conv block num of each pan block
            drop_block (bool): whether use DropBlock or not
            block_size (int): block size of DropBlock
            keep_prob (float): keep probability of DropBlock
            spp (bool): whether use spp or not

        """
        super(PPYOLOFPN, self).__init__()
        assert len(in_channels) > 0, "in_channels length should > 0"
        self.in_channels = in_channels
        self.num_blocks = len(in_channels)
        # parse kwargs
        self.coord_conv = coord_conv
        self.drop_block = drop_block
        self.block_size = block_size
        self.keep_prob = keep_prob
        self.spp = spp
        self.conv_block_num = conv_block_num
        self.data_format = data_format
        if self.coord_conv:
            ConvLayer = CoordConv
        else:
            ConvLayer = ConvBNLayer

        if self.drop_block:
            dropblock_cfg = [[
                'dropblock', DropBlock, [self.block_size, self.keep_prob],
                dict()
            ]]
        else:
            dropblock_cfg = []

        self._out_channels = []
        self.yolo_blocks = []
        self.routes = []
        for i, ch_in in enumerate(self.in_channels[::-1]):
            if i > 0:
                ch_in += 512 // (2**i)
            channel = 64 * (2**self.num_blocks) // (2**i)
            base_cfg = []
            c_in, c_out = ch_in, channel
            for j in range(self.conv_block_num):
                base_cfg += [
                    [
                        'conv{}'.format(2 * j), ConvLayer, [c_in, c_out, 1],
                        dict(
                            padding=0, norm_type=norm_type)
                    ],
                    [
                        'conv{}'.format(2 * j + 1), ConvBNLayer,
                        [c_out, c_out * 2, 3], dict(
                            padding=1, norm_type=norm_type)
                    ],
                ]
                c_in, c_out = c_out * 2, c_out

            base_cfg += [[
                'route', ConvLayer, [c_in, c_out, 1], dict(
                    padding=0, norm_type=norm_type)
            ], [
                'tip', ConvLayer, [c_out, c_out * 2, 3], dict(
                    padding=1, norm_type=norm_type)
            ]]

            if self.conv_block_num == 2:
                if i == 0:
                    if self.spp:
                        spp_cfg = [[
                            'spp', SPP, [channel * 4, channel, 1], dict(
                                pool_size=[5, 9, 13], norm_type=norm_type)
                        ]]
                    else:
                        spp_cfg = []
                    cfg = base_cfg[0:3] + spp_cfg + base_cfg[
                        3:4] + dropblock_cfg + base_cfg[4:6]
                else:
                    cfg = base_cfg[0:2] + dropblock_cfg + base_cfg[2:6]
            elif self.conv_block_num == 0:
                if self.spp and i == 0:
                    spp_cfg = [[
                        'spp', SPP, [c_in * 4, c_in, 1], dict(
                            pool_size=[5, 9, 13], norm_type=norm_type)
                    ]]
                else:
                    spp_cfg = []
                cfg = spp_cfg + dropblock_cfg + base_cfg
            name = 'yolo_block.{}'.format(i)
            yolo_block = self.add_sublayer(name, PPYOLODetBlock(cfg, name))
            self.yolo_blocks.append(yolo_block)
            self._out_channels.append(channel * 2)
            if i < self.num_blocks - 1:
                name = 'yolo_transition.{}'.format(i)
                route = self.add_sublayer(
                    name,
                    ConvBNLayer(
                        ch_in=channel,
                        ch_out=256 // (2**i),
                        filter_size=1,
                        stride=1,
                        padding=0,
                        norm_type=norm_type,
                        data_format=data_format,
                        name=name))
                self.routes.append(route)

    def forward(self, blocks):
        assert len(blocks) == self.num_blocks
        blocks = blocks[::-1]
        yolo_feats = []
        for i, block in enumerate(blocks):
            if i > 0:
                if self.data_format == 'NCHW':
                    logger.info("进入ChannelA")
                    block = ChannelAttention(block)
                    block = SpatialAttention()
                    block = paddle.concat([route, block], axis=1)
                else:
                    block = ChannelAttention(block)
                    block = SpatialAttention()
                    block = paddle.concat([route, block], axis=-1)
            route, tip = self.yolo_blocks[i](block)
            yolo_feats.append(tip)

            if i < self.num_blocks - 1:
                route = self.routes[i](route)
                route = F.interpolate(
                    route, scale_factor=2., data_format=self.data_format)

        return yolo_feats

    @classmethod
    def from_config(cls, cfg, input_shape):
        return {'in_channels': [i.channels for i in input_shape], }

    @property
    def out_shape(self):
        return [ShapeSpec(channels=c) for c in self._out_channels]

class ChannelAttention(nn.Layer):
    def __init__(self, in_planes, ratio=16):
        super(ChannelAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2D(1)
        self.max_pool = nn.AdaptiveAvgPool2D(1)

        self.fc1   = nn.Conv2D(in_planes, in_planes // 16, 1, bias=False)
        self.relu1 = F.relu()
        self.fc2   = nn.Conv2D(in_planes // 16, in_planes, 1, bias=False)

        self.sigmoid = F.sigmoid()

    def forward(self, x):
        logger.info("进入ChannelAttention")
        
        avg_out = self.fc2(self.relu1(self.fc1(self.avg_pool(x))))
        max_out = self.fc2(self.relu1(self.fc1(self.max_pool(x))))
        out = avg_out + max_out
        return self.sigmoid(out)

class SpatialAttention(nn.Layer):
    def __init__(self, kernel_size=7):
        super(SpatialAttention, self).__init__()
        logger.info("进入SpatialAttention")

        assert kernel_size in (3, 7), 'kernel size must be 3 or 7'
        padding = 3 if kernel_size == 7 else 1

        self.conv1 = nn.Conv2D(2, 1, kernel_size, padding=padding, bias=False)
        self.sigmoid = F.sigmoid()

    def forward(self, x):
        avg_out = paddle.mean(x, dim=1, keepdim=True)
        max_out, _ = paddle.max(x, dim=1, keepdim=True)
        x = paddle.concat([avg_out, max_out], dim=1)
        x = self.conv1(x)
        return self.sigmoid(x)

paddlepaddle - gpu 版本2.0.0 PaddleDetection版本release 2.0 rc 硬件：jetson nano jetpack4.3 场景： 1、使用tools/train.py 对 路标检测模型（yolov3_mobilenet_v1_roadsign.yml）进行训练，得到best_model。 2、使用tools/export_model.py进行导出模型。 python tools/export_model.py -c configs/yolov3_mobilenet_v1_roadsign.yml
--output_dir=./export_model
-o weights=./output/yolov3_mobilenet_v1_roadsign/best_model TestReader,input_def.image_shape=[3,320,320] 3、使用deploy/python/infer.py在jetson nano下进行部署。 （1）使用gpu预测： (2)使用cpu预测正常显示。

麻烦跟进一下，后续需要提供什么信息请告知我就行，谢谢！

文档链接&描述 Document Links & Description

链接：https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/visdrone 第一段存在问题：PaddleDetection团队提供了针对VisDrone-DET小目标数航拍场景的基于PP-YOLOE的检测模型，用户可以下载模型进行使用。整理后的COCO格式VisDrone-DET数据集下载链接，检测其中的10类，包括 pedestrian(1), people(2), bicycle(3), car(4), van(5), truck(6), tricycle(7), awning-tricycle(8), bus(9), motor(10)，原始数据集下载链接。 问题：这一部分“检测其中的10类，包括 pedestrian(1), people(2), bicycle(3), car(4), van(5), truck(6), tricycle(7), awning-tricycle(8), bus(9), motor(10)。”类别名和序号应该存在问题。 发现问题步骤： 1、执行如下命令python tools/infer.py -c configs/smalldet/ppyoloe_crn_l_80e_sliced_visdrone_640_025.yml --infer_img=demo/000000014439.jpg -o use_gpu=False weights=https://paddledet.bj.bcebos.com/models/ppyoloe_crn_l_80e_sliced_visdrone_640_025.pdparams --infer_img=demo/000000014439.jpg --draw_threshold=0.1 2、output目录生成的文件打印的类别中包括traffic_light。

请提出你的建议 Please give your suggestion

从结果来看应该是有错的，暂时我也没找到正确的类别。

问题确认 Search before asking

• [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

为什么使用多进程以后，时间会更慢了？本来行人检测时间也就6ms左右，用了如下方式启用多进程以后，该行人检测时间竟然达到了48ms以上，这是为什么？ read_frame函数在做读帧，并做行人检测处理。

问题确认 Search before asking

• [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

当用paddledetection中的车牌检测时。车头有文字的时候，没有检测到车牌，而是检测到文字上了，想问问能怎么解决!

问题确认 Search before asking

• [X] 我已经查询历史issue，没有发现相似的bug。I have searched the issues and found no similar bug report.

Bug组件 Bug Component

No response

Bug描述 Describe the Bug

多进程代码写法是参考网上的，能跑得通，但是用在飞桨这里就报错 报错如下：

复现环境 Environment

linux:PadlepadleDetection

Bug描述确认 Bug description confirmation

• [X] 我确认已经提供了Bug复现步骤、代码改动说明、以及环境信息，确认问题是可以复现的。I confirm that the bug replication steps, code change instructions, and environment information have been provided, and the problem can be reproduced.

是否愿意提交PR？ Are you willing to submit a PR?

• [X] 我愿意提交PR！I'd like to help by submitting a PR!