微信公众号 讲师中心
首页
文章
后端开发 web前端 数据库 开发工具 php框架 常见问题 科技 Java 系统教程 电脑教程 硬件教程 手机教程 软件教程 游戏教程 自媒体 新闻
专题
后端开发 web前端 数据库 开发工具 php框架 科技 Java 系统教程 电脑教程 硬件教程 手机教程 软件教程 游戏教程 新闻
AI工具
AI 聊天问答 Agent智能体 AI 文本写作 AI 绘画作图 AI 设计工具 AI 视频创作 AI 音频制作 AI 办公学习 AI 编程开发 Prompt指令
学习
大前端 后端开发 数据库 移动端 运维开发 计算机基础
编程手册
大前端 后端开发 数据库 移动端 运维开发 计算机基础
下载
js特效 网站源码 工具下载 类库下载 网站素材 学习资源 插件扩展 手机/移动开发 手机游戏
搜索
后端开发 web前端 数据库 开发工具 php框架 常见问题 科技 Java 系统教程 电脑教程 硬件教程 手机教程 软件教程 游戏教程
自媒体 新闻
首页 > 科技周边 > 人工智能 > 正文

基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例

P粉084495128
发布： 2025-07-23 11:08:23
原创
830人浏览过
本文介绍基于PaddleSeg 2.1实现DLRSD遥感影像数据集语义分割的流程。先准备数据集，按自定义格式整理，划分训练、验证集。接着用OCRNet模型训练，迭代1000次，最佳mIoU达0.6823。测试采用增强模式，mIoU为0.6839，最后完成单张影像推理，展示结果。

☞☞☞AI 智能聊天, 问答助手, AI 智能搜索, 免费无限量使用 DeepSeek R1 模型☜☜☜

基于paddleseg的自定义遥感数据语义分割——以dlrsd为例 - php中文网

基于paddleSeg进行自定义遥感影像数据集的语义分割——以DLRSD为例

  • 基于paddleseg 2.1
  • 使用自定义数据集DLRSD,其他遥感数据集
  • 实现训练、测试、推理
  • 脚本版任务

数据格式准备

DLRSD数据集

  • 基于UCMerced_LandUse数据集进行标注
  • 标注了17种类型的遥感地物类型
  • 图像尺寸256 * 256 * 3
  • 图像数21* 100 = 2100
  • gt格式，单通道8bit彩图，推荐使用PIL读取
  • DLRSD详细数据信息

数据展示： 基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例 - php中文网 基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例 - php中文网        

PaddleSeg数据格式

paddleSeg支持CityScapes、ADE20K、Pascal VOC、自定义等数据集格式，paddleseg数据集配置链接

       

本文使用的是自定义数据格式：

  • 数据存放在PaddleSeg/dataset/dlrsd文件夹下，
  • 影像、标注分别存储到image、gt下，
  • train、val、test等部分按txt分别存储

               

文件结构：

    custom_dataset
        |
        |--images        |  |--image1.jpg        |  |--image2.jpg        |  |--...        |
        |--labels        |  |--label1.jpg        |  |--label2.png        |  |--...        |
        |--train.txt        |
        |--val.txt        |
        |--test.txt
登录后复制
       

txt内容格式

images/image1.jpg labels/label1.png
images/image2.jpg labels/label2.png
...
登录后复制
       

数据集配置文件可以参考设置如下：

train_dataset:
  type: Dataset
  dataset_root: custom_dataset
  train_path: custom_dataset/train.txt
  num_classes: 2
  transforms:
    - type: ResizeStepScaling
      min_scale_factor: 0.5
      max_scale_factor: 2.0
      scale_step_size: 0.25
    - type: RandomPaddingCrop
      crop_size: [512, 512]    - type: RandomHorizontalFlip
    - type: Normalize
  mode: train
登录后复制
   

项目准备

In [ ]
# paddleseg下载解压# 从gitee下载PaddleSeg!git clone https://gitee.com/paddlepaddle/PaddleSeg.git# 进入PaddleSeg目录%cd PaddleSeg# 切换到release/2.1分支!git checkout -b release/2.1 origin/release/2.1# 安装依赖!pip install -r requirements.txt
登录后复制
       
fatal: destination path 'PaddleSeg' already exists and is not an empty directory.
/home/aistudio/PaddleSeg
fatal: Not a git repository (or any parent up to mount point /home/aistudio)
Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set).
Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple
Requirement already satisfied: pre-commit in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 1)) (1.21.0)
Requirement already satisfied: yapf==0.26.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 2)) (0.26.0)
Requirement already satisfied: flake8 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 3)) (4.0.1)
Requirement already satisfied: pyyaml>=5.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 4)) (5.1.2)
Requirement already satisfied: visualdl>=2.0.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 5)) (2.2.0)
Requirement already satisfied: opencv-python in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 6)) (4.1.1.26)
Requirement already satisfied: tqdm in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 7)) (4.36.1)
Requirement already satisfied: filelock in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 8)) (3.0.12)
Requirement already satisfied: scipy in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 9)) (1.6.3)
Requirement already satisfied: prettytable in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from -r requirements.txt (line 10)) (0.7.2)
Requirement already satisfied: six in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (1.16.0)
Requirement already satisfied: toml in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (0.10.0)
Requirement already satisfied: nodeenv>=0.11.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (1.3.4)
Requirement already satisfied: virtualenv>=15.2 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (16.7.9)
Requirement already satisfied: identify>=1.0.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (1.4.10)
Requirement already satisfied: importlib-metadata in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (4.2.0)
Requirement already satisfied: cfgv>=2.0.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (2.0.1)
Requirement already satisfied: aspy.yaml in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from pre-commit->-r requirements.txt (line 1)) (1.3.0)
Requirement already satisfied: pycodestyle<2.9.0,>=2.8.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flake8->-r requirements.txt (line 3)) (2.8.0)
Requirement already satisfied: mccabe<0.7.0,>=0.6.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flake8->-r requirements.txt (line 3)) (0.6.1)
Requirement already satisfied: pyflakes<2.5.0,>=2.4.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flake8->-r requirements.txt (line 3)) (2.4.0)
Requirement already satisfied: flask>=1.1.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (1.1.1)
Requirement already satisfied: requests in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (2.27.1)
Requirement already satisfied: shellcheck-py in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (0.7.1.1)
Requirement already satisfied: bce-python-sdk in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (0.8.53)
Requirement already satisfied: protobuf>=3.11.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (3.14.0)
Requirement already satisfied: Flask-Babel>=1.0.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (1.0.0)
Requirement already satisfied: numpy in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (1.20.3)
Requirement already satisfied: matplotlib in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (2.2.3)
Requirement already satisfied: Pillow>=7.0.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (7.1.2)
Requirement already satisfied: pandas in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from visualdl>=2.0.0->-r requirements.txt (line 5)) (1.1.5)
Requirement already satisfied: Werkzeug>=0.15 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flask>=1.1.1->visualdl>=2.0.0->-r requirements.txt (line 5)) (0.16.0)
Requirement already satisfied: itsdangerous>=0.24 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flask>=1.1.1->visualdl>=2.0.0->-r requirements.txt (line 5)) (1.1.0)
Requirement already satisfied: Jinja2>=2.10.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flask>=1.1.1->visualdl>=2.0.0->-r requirements.txt (line 5)) (3.0.0)
Requirement already satisfied: click>=5.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from flask>=1.1.1->visualdl>=2.0.0->-r requirements.txt (line 5)) (7.0)
Requirement already satisfied: pytz in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from Flask-Babel>=1.0.0->visualdl>=2.0.0->-r requirements.txt (line 5)) (2022.1)
Requirement already satisfied: Babel>=2.3 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from Flask-Babel>=1.0.0->visualdl>=2.0.0->-r requirements.txt (line 5)) (2.9.1)
Requirement already satisfied: typing-extensions>=3.6.4 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from importlib-metadata->pre-commit->-r requirements.txt (line 1)) (4.1.1)
Requirement already satisfied: zipp>=0.5 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from importlib-metadata->pre-commit->-r requirements.txt (line 1)) (3.7.0)
Requirement already satisfied: pycryptodome>=3.8.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from bce-python-sdk->visualdl>=2.0.0->-r requirements.txt (line 5)) (3.9.9)
Requirement already satisfied: future>=0.6.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from bce-python-sdk->visualdl>=2.0.0->-r requirements.txt (line 5)) (0.18.0)
Requirement already satisfied: python-dateutil>=2.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from matplotlib->visualdl>=2.0.0->-r requirements.txt (line 5)) (2.8.2)
Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from matplotlib->visualdl>=2.0.0->-r requirements.txt (line 5)) (3.0.7)
Requirement already satisfied: cycler>=0.10 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from matplotlib->visualdl>=2.0.0->-r requirements.txt (line 5)) (0.10.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from matplotlib->visualdl>=2.0.0->-r requirements.txt (line 5)) (1.1.0)
Requirement already satisfied: certifi>=2017.4.17 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from requests->visualdl>=2.0.0->-r requirements.txt (line 5)) (2021.10.8)
Requirement already satisfied: idna<4,>=2.5 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from requests->visualdl>=2.0.0->-r requirements.txt (line 5)) (3.3)
Requirement already satisfied: charset-normalizer~=2.0.0 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from requests->visualdl>=2.0.0->-r requirements.txt (line 5)) (2.0.12)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from requests->visualdl>=2.0.0->-r requirements.txt (line 5)) (1.26.9)
Requirement already satisfied: MarkupSafe>=2.0.0rc2 in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from Jinja2>=2.10.1->flask>=1.1.1->visualdl>=2.0.0->-r requirements.txt (line 5)) (2.0.1)
Requirement already satisfied: setuptools in /opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages (from kiwisolver>=1.0.1->matplotlib->visualdl>=2.0.0->-r requirements.txt (line 5)) (56.2.0)
登录后复制
       

预修改的文件：

为正常运行网络模型，本项目已对以下文件进行修改，详情查看代码

硅基智能
硅基智能

基于Web3.0的元宇宙，去中心化的互联网，高质量、沉浸式元宇宙直播平台，用数字化重新定义直播

硅基智能 62
查看详情 硅基智能
  • PaddleSeg/configs/base/dlrsd.yml——(数据集配置文件)
  • PaddleSeg/configs/unet/unet_dlrds.yml——(模型配置文件)
  • PaddleSeg/ocrnet_dlrsd.yml——(ocrnet模型配置文件)
  • work/img_copy.py——(文件拷贝代码)
  • work/data_split.py——(数据划分代码)
  • PaddleSeg/predict.py——(增加tif读取)

数据准备

In [ ]
# 文件解压!unzip -oq /home/aistudio/data/data55005/UCMerced_LandUse.zip -d /home/aistudio/data/image
!unzip -oq /home/aistudio/data/data55005/DLRSD.zip  -d /home/aistudio/data/gt_color
登录后复制
   
In [ ]
# 数据预处理，将文件组织自定义数据集格式# image copy!python /home/aistudio/work/img_copy.py /home/aistudio/data/image/UCMerced_LandUse/Images /home/aistudio/PaddleSeg/dataset/dlrsd/image# gt copy!python /home/aistudio/work/img_copy.py /home/aistudio/data/gt_color/DLRSD/Images /home/aistudio/PaddleSeg/dataset/dlrsd/gt# dataset split!python /home/aistudio/work/data_split.py 0.8 0.2 0 /home/aistudio/PaddleSeg/dataset/dlrsd/image /home/aistudio/PaddleSeg/dataset/dlrsd/gt
登录后复制
       
copy from /home/aistudio/data/image/UCMerced_LandUse/Images to /home/aistudio/PaddleSeg/dataset/dlrsd/image:
agricultural finish
airplane finish
tenniscourt finish
chaparral finish
freeway finish
parkinglot finish
mobilehomepark finish
buildings finish
sparseresidential finish
baseballdiamond finish
runway finish
denseresidential finish
beach finish
storagetanks finish
river finish
forest finish
overpass finish
golfcourse finish
harbor finish
intersection finish
mediumresidential finish
copy from /home/aistudio/data/gt_color/DLRSD/Images to /home/aistudio/PaddleSeg/dataset/dlrsd/gt:
agricultural finish
airplane finish
tenniscourt finish
chaparral finish
freeway finish
parkinglot finish
mobilehomepark finish
buildings finish
sparseresidential finish
baseballdiamond finish
runway finish
denseresidential finish
beach finish
storagetanks finish
river finish
forest finish
overpass finish
golfcourse finish
harbor finish
intersection finish
mediumresidential finish
total 2100, split: train 1680 - 0.80 rate, val 420 - 0.20 rate, test 0 - 0.00 rate
登录后复制
       

模型训练

参考paddleSeg

In [8]
# 模型训练, 其他超参数可参考paddleSeg，对unet_dlrds.yml进行修改# unet_dlrds.yml中的参数会覆盖dlrsd.yml中的参数，修改参数时，可以直接在unet_dlrds.yml中添加参数# !python train.py \#        --config configs/unet/unet_dlrsd.yml \#        --do_eval \#        --use_vdl \#        --save_interval 100 \#        --save_dir output# 使用精度更高的ocrnet进行训练, 可以尝试10000以上iter进行训练，得到更高精度!python train.py \
       --batch_size 32 \
       --iters 1000 \
       --config ocrnet_dlrsd.yml \
       --do_eval \
       --use_vdl \
       --save_interval 100 \
       --save_dir output
登录后复制
       
2022-04-14 12:18:35 [INFO]
------------Environment Information-------------
platform: Linux-4.13.0-36-generic-x86_64-with-debian-stretch-sid
Python: 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0]
Paddle compiled with cuda: True
NVCC: Cuda compilation tools, release 10.1, V10.1.243
cudnn: 7.6
GPUs used: 1
CUDA_VISIBLE_DEVICES: None
GPU: ['GPU 0: Tesla V100-SXM2-16GB']
GCC: gcc (Ubuntu 7.5.0-3ubuntu1~16.04) 7.5.0
PaddlePaddle: 2.1.2
OpenCV: 4.1.1
------------------------------------------------
2022-04-14 12:18:35 [INFO]
---------------Config Information---------------
batch_size: 32
iters: 1000
loss:
  coef:
  - 1
  - 0.4
  types:
  - ignore_index: 255
    type: CrossEntropyLoss
  - ignore_index: 255
    type: CrossEntropyLoss
lr_scheduler:
  end_lr: 0
  learning_rate: 0.01
  power: 0.9
  type: PolynomialDecay
model:
  backbone:
    pretrained: https://bj.bcebos.com/paddleseg/dygraph/hrnet_w18_ssld.tar.gz
    type: HRNet_W18
  backbone_indices:
  - 0
  num_classes: 18
  type: OCRNet
optimizer:
  momentum: 0.9
  type: sgd
  weight_decay: 4.0e-05
train_dataset:
  dataset_root: dataset/dlrsd
  mode: train
  num_classes: 18
  train_path: dataset/dlrsd/train.txt
  transforms:
  - max_scale_factor: 2.0
    min_scale_factor: 0.5
    scale_step_size: 0.25
    type: ResizeStepScaling
  - crop_size:
    - 256
    - 256
    type: RandomPaddingCrop
  - type: RandomHorizontalFlip
  - type: Normalize
  type: Dataset
val_dataset:
  dataset_root: dataset/dlrsd
  mode: val
  num_classes: 18
  transforms:
  - type: Normalize
  - target_size:
    - 256
    - 256
    type: Resize
  type: Dataset
  val_path: dataset/dlrsd/val.txt
------------------------------------------------
W0414 12:18:35.587487   602 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.0, Runtime API Version: 10.1
W0414 12:18:35.587553   602 device_context.cc:422] device: 0, cuDNN Version: 7.6.
2022-04-14 12:18:39 [INFO]	Loading pretrained model from https://bj.bcebos.com/paddleseg/dygraph/hrnet_w18_ssld.tar.gz
2022-04-14 12:18:40 [INFO]	There are 1525/1525 variables loaded into HRNet.
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/framework.py:706: DeprecationWarning: `np.bool` is a deprecated alias for the builtin `bool`. To silence this warning, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.
Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
  elif dtype == np.bool:
2022-04-14 12:18:47 [INFO]	[TRAIN] epoch: 1, iter: 10/1000, loss: 3.4345, lr: 0.009919, batch_cost: 0.6647, reader_cost: 0.03684, ips: 48.1407 samples/sec | ETA 00:10:58
2022-04-14 12:18:53 [INFO]	[TRAIN] epoch: 1, iter: 20/1000, loss: 2.7537, lr: 0.009829, batch_cost: 0.6197, reader_cost: 0.00049, ips: 51.6372 samples/sec | ETA 00:10:07
2022-04-14 12:18:59 [INFO]	[TRAIN] epoch: 1, iter: 30/1000, loss: 2.4857, lr: 0.009739, batch_cost: 0.5961, reader_cost: 0.00085, ips: 53.6849 samples/sec | ETA 00:09:38
2022-04-14 12:19:05 [INFO]	[TRAIN] epoch: 1, iter: 40/1000, loss: 2.2253, lr: 0.009648, batch_cost: 0.6049, reader_cost: 0.00106, ips: 52.8983 samples/sec | ETA 00:09:40
2022-04-14 12:19:11 [INFO]	[TRAIN] epoch: 1, iter: 50/1000, loss: 1.9626, lr: 0.009558, batch_cost: 0.5984, reader_cost: 0.00085, ips: 53.4791 samples/sec | ETA 00:09:28
2022-04-14 12:19:17 [INFO]	[TRAIN] epoch: 2, iter: 60/1000, loss: 1.8616, lr: 0.009467, batch_cost: 0.6030, reader_cost: 0.01686, ips: 53.0698 samples/sec | ETA 00:09:26
2022-04-14 12:19:23 [INFO]	[TRAIN] epoch: 2, iter: 70/1000, loss: 1.7680, lr: 0.009377, batch_cost: 0.5849, reader_cost: 0.00077, ips: 54.7057 samples/sec | ETA 00:09:04
2022-04-14 12:19:28 [INFO]	[TRAIN] epoch: 2, iter: 80/1000, loss: 1.6116, lr: 0.009286, batch_cost: 0.5890, reader_cost: 0.00095, ips: 54.3256 samples/sec | ETA 00:09:01
2022-04-14 12:19:35 [INFO]	[TRAIN] epoch: 2, iter: 90/1000, loss: 1.5829, lr: 0.009195, batch_cost: 0.6182, reader_cost: 0.00047, ips: 51.7606 samples/sec | ETA 00:09:22
2022-04-14 12:19:41 [INFO]	[TRAIN] epoch: 2, iter: 100/1000, loss: 1.5083, lr: 0.009104, batch_cost: 0.5941, reader_cost: 0.00166, ips: 53.8610 samples/sec | ETA 00:08:54
2022-04-14 12:19:41 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/dygraph/math_op_patch.py:239: UserWarning: The dtype of left and right variables are not the same, left dtype is paddle.int32, but right dtype is paddle.bool, the right dtype will convert to paddle.int32
  format(lhs_dtype, rhs_dtype, lhs_dtype))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/dygraph/math_op_patch.py:239: UserWarning: The dtype of left and right variables are not the same, left dtype is paddle.int64, but right dtype is paddle.bool, the right dtype will convert to paddle.int64
  format(lhs_dtype, rhs_dtype, lhs_dtype))
420/420 [==============================] - 60s 142ms/step - batch_cost: 0.1419 - reader cost: 3.9444e-
2022-04-14 12:20:40 [INFO]	[EVAL] #Images: 420 mIoU: 0.3062 Acc: 0.6152 Kappa: 0.5491 
2022-04-14 12:20:40 [INFO]	[EVAL] Class IoU: 
[0.     0.     0.2566 0.1382 0.4141 0.4006 0.3423 0.     0.9608 0.4201
 0.0166 0.6178 0.2826 0.5871 0.15   0.     0.6299 0.294 ]
2022-04-14 12:20:40 [INFO]	[EVAL] Class Acc: 
[0.     0.     0.4912 0.9435 0.4735 0.4704 0.3853 0.     1.     0.8808
 0.086  0.6741 0.5215 0.85   0.1504 0.     0.6799 0.3236]
2022-04-14 12:20:41 [INFO]	[EVAL] The model with the best validation mIoU (0.3062) was saved at iter 100.
2022-04-14 12:20:47 [INFO]	[TRAIN] epoch: 3, iter: 110/1000, loss: 1.3279, lr: 0.009013, batch_cost: 0.5924, reader_cost: 0.01635, ips: 54.0138 samples/sec | ETA 00:08:47
2022-04-14 12:20:53 [INFO]	[TRAIN] epoch: 3, iter: 120/1000, loss: 1.2831, lr: 0.008922, batch_cost: 0.5966, reader_cost: 0.00139, ips: 53.6371 samples/sec | ETA 00:08:45
2022-04-14 12:20:59 [INFO]	[TRAIN] epoch: 3, iter: 130/1000, loss: 1.1546, lr: 0.008831, batch_cost: 0.6096, reader_cost: 0.00082, ips: 52.4938 samples/sec | ETA 00:08:50
2022-04-14 12:21:05 [INFO]	[TRAIN] epoch: 3, iter: 140/1000, loss: 1.2061, lr: 0.008740, batch_cost: 0.5993, reader_cost: 0.00117, ips: 53.3943 samples/sec | ETA 00:08:35
2022-04-14 12:21:11 [INFO]	[TRAIN] epoch: 3, iter: 150/1000, loss: 1.0794, lr: 0.008648, batch_cost: 0.5980, reader_cost: 0.00177, ips: 53.5158 samples/sec | ETA 00:08:28
2022-04-14 12:21:17 [INFO]	[TRAIN] epoch: 4, iter: 160/1000, loss: 1.0784, lr: 0.008557, batch_cost: 0.5858, reader_cost: 0.01779, ips: 54.6240 samples/sec | ETA 00:08:12
2022-04-14 12:21:23 [INFO]	[TRAIN] epoch: 4, iter: 170/1000, loss: 1.0746, lr: 0.008465, batch_cost: 0.5943, reader_cost: 0.00015, ips: 53.8473 samples/sec | ETA 00:08:13
2022-04-14 12:21:29 [INFO]	[TRAIN] epoch: 4, iter: 180/1000, loss: 1.0160, lr: 0.008374, batch_cost: 0.5921, reader_cost: 0.00016, ips: 54.0444 samples/sec | ETA 00:08:05
2022-04-14 12:21:35 [INFO]	[TRAIN] epoch: 4, iter: 190/1000, loss: 1.1226, lr: 0.008282, batch_cost: 0.5955, reader_cost: 0.00195, ips: 53.7374 samples/sec | ETA 00:08:02
2022-04-14 12:21:41 [INFO]	[TRAIN] epoch: 4, iter: 200/1000, loss: 1.1066, lr: 0.008190, batch_cost: 0.6310, reader_cost: 0.00169, ips: 50.7141 samples/sec | ETA 00:08:24
2022-04-14 12:21:41 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 61s 144ms/step - batch_cost: 0.1440 - reader cost: 4.2161e-0
2022-04-14 12:22:41 [INFO]	[EVAL] #Images: 420 mIoU: 0.5294 Acc: 0.7634 Kappa: 0.7237 
2022-04-14 12:22:41 [INFO]	[EVAL] Class IoU: 
[0.     0.4053 0.4327 0.627  0.599  0.0673 0.7334 0.     0.9533 0.5389
 0.5181 0.7464 0.4466 0.8169 0.552  0.6634 0.69   0.7388]
2022-04-14 12:22:41 [INFO]	[EVAL] Class Acc: 
[0.     0.6584 0.5536 0.7176 0.7343 0.8204 0.7856 0.     0.9868 0.6825
 0.7245 0.8834 0.8107 0.9022 0.5718 0.8478 0.7959 0.9131]
2022-04-14 12:22:42 [INFO]	[EVAL] The model with the best validation mIoU (0.5294) was saved at iter 200.
2022-04-14 12:22:48 [INFO]	[TRAIN] epoch: 5, iter: 210/1000, loss: 1.0923, lr: 0.008098, batch_cost: 0.5964, reader_cost: 0.01708, ips: 53.6531 samples/sec | ETA 00:07:51
2022-04-14 12:22:54 [INFO]	[TRAIN] epoch: 5, iter: 220/1000, loss: 1.0190, lr: 0.008005, batch_cost: 0.6013, reader_cost: 0.00018, ips: 53.2211 samples/sec | ETA 00:07:48
2022-04-14 12:23:00 [INFO]	[TRAIN] epoch: 5, iter: 230/1000, loss: 1.0221, lr: 0.007913, batch_cost: 0.5972, reader_cost: 0.00066, ips: 53.5837 samples/sec | ETA 00:07:39
2022-04-14 12:23:06 [INFO]	[TRAIN] epoch: 5, iter: 240/1000, loss: 0.9766, lr: 0.007821, batch_cost: 0.6007, reader_cost: 0.00062, ips: 53.2710 samples/sec | ETA 00:07:36
2022-04-14 12:23:12 [INFO]	[TRAIN] epoch: 5, iter: 250/1000, loss: 0.8989, lr: 0.007728, batch_cost: 0.6137, reader_cost: 0.00145, ips: 52.1419 samples/sec | ETA 00:07:40
2022-04-14 12:23:18 [INFO]	[TRAIN] epoch: 5, iter: 260/1000, loss: 0.9029, lr: 0.007635, batch_cost: 0.5650, reader_cost: 0.00090, ips: 56.6328 samples/sec | ETA 00:06:58
2022-04-14 12:23:24 [INFO]	[TRAIN] epoch: 6, iter: 270/1000, loss: 0.8537, lr: 0.007543, batch_cost: 0.6184, reader_cost: 0.01685, ips: 51.7439 samples/sec | ETA 00:07:31
2022-04-14 12:23:30 [INFO]	[TRAIN] epoch: 6, iter: 280/1000, loss: 0.8544, lr: 0.007450, batch_cost: 0.5992, reader_cost: 0.00116, ips: 53.4018 samples/sec | ETA 00:07:11
2022-04-14 12:23:36 [INFO]	[TRAIN] epoch: 6, iter: 290/1000, loss: 0.8475, lr: 0.007357, batch_cost: 0.5923, reader_cost: 0.00041, ips: 54.0301 samples/sec | ETA 00:07:00
2022-04-14 12:23:42 [INFO]	[TRAIN] epoch: 6, iter: 300/1000, loss: 0.7936, lr: 0.007264, batch_cost: 0.6050, reader_cost: 0.00058, ips: 52.8951 samples/sec | ETA 00:07:03
2022-04-14 12:23:42 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 144ms/step - batch_cost: 0.1435 - reader cost: 4.1186e-0
2022-04-14 12:24:42 [INFO]	[EVAL] #Images: 420 mIoU: 0.5801 Acc: 0.7912 Kappa: 0.7568 
2022-04-14 12:24:42 [INFO]	[EVAL] Class IoU: 
[0.     0.5324 0.4741 0.626  0.6086 0.5252 0.7589 0.007  0.9623 0.5803
 0.491  0.7786 0.5816 0.9076 0.5717 0.616  0.6964 0.7239]
2022-04-14 12:24:42 [INFO]	[EVAL] Class Acc: 
[0.     0.6716 0.7091 0.8305 0.8657 0.7009 0.8022 0.9908 0.9958 0.7105
 0.5133 0.8871 0.7654 0.9173 0.6074 0.6626 0.7855 0.9062]
2022-04-14 12:24:43 [INFO]	[EVAL] The model with the best validation mIoU (0.5801) was saved at iter 300.
2022-04-14 12:24:49 [INFO]	[TRAIN] epoch: 6, iter: 310/1000, loss: 0.8334, lr: 0.007170, batch_cost: 0.6216, reader_cost: 0.00088, ips: 51.4776 samples/sec | ETA 00:07:08
2022-04-14 12:24:55 [INFO]	[TRAIN] epoch: 7, iter: 320/1000, loss: 0.8206, lr: 0.007077, batch_cost: 0.6054, reader_cost: 0.01669, ips: 52.8564 samples/sec | ETA 00:06:51
2022-04-14 12:25:02 [INFO]	[TRAIN] epoch: 7, iter: 330/1000, loss: 0.7791, lr: 0.006983, batch_cost: 0.6115, reader_cost: 0.00038, ips: 52.3275 samples/sec | ETA 00:06:49
2022-04-14 12:25:08 [INFO]	[TRAIN] epoch: 7, iter: 340/1000, loss: 0.7893, lr: 0.006889, batch_cost: 0.5994, reader_cost: 0.00075, ips: 53.3874 samples/sec | ETA 00:06:35
2022-04-14 12:25:14 [INFO]	[TRAIN] epoch: 7, iter: 350/1000, loss: 0.7701, lr: 0.006796, batch_cost: 0.5998, reader_cost: 0.00048, ips: 53.3533 samples/sec | ETA 00:06:29
2022-04-14 12:25:19 [INFO]	[TRAIN] epoch: 7, iter: 360/1000, loss: 0.7948, lr: 0.006702, batch_cost: 0.5912, reader_cost: 0.00020, ips: 54.1302 samples/sec | ETA 00:06:18
2022-04-14 12:25:25 [INFO]	[TRAIN] epoch: 8, iter: 370/1000, loss: 0.7359, lr: 0.006607, batch_cost: 0.5961, reader_cost: 0.01591, ips: 53.6839 samples/sec | ETA 00:06:15
2022-04-14 12:25:31 [INFO]	[TRAIN] epoch: 8, iter: 380/1000, loss: 0.7732, lr: 0.006513, batch_cost: 0.5958, reader_cost: 0.00133, ips: 53.7118 samples/sec | ETA 00:06:09
2022-04-14 12:25:37 [INFO]	[TRAIN] epoch: 8, iter: 390/1000, loss: 0.7243, lr: 0.006419, batch_cost: 0.6008, reader_cost: 0.00178, ips: 53.2630 samples/sec | ETA 00:06:06
2022-04-14 12:25:43 [INFO]	[TRAIN] epoch: 8, iter: 400/1000, loss: 0.7617, lr: 0.006324, batch_cost: 0.5974, reader_cost: 0.00041, ips: 53.5635 samples/sec | ETA 00:05:58
2022-04-14 12:25:43 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 142ms/step - batch_cost: 0.1421 - reader cost: 3.9464e-0
2022-04-14 12:26:43 [INFO]	[EVAL] #Images: 420 mIoU: 0.6354 Acc: 0.8046 Kappa: 0.7733 
2022-04-14 12:26:43 [INFO]	[EVAL] Class IoU: 
[0.     0.5345 0.4958 0.6375 0.6488 0.5401 0.7947 0.4412 0.9566 0.6051
 0.5929 0.7681 0.6005 0.932  0.6718 0.6875 0.7183 0.812 ]
2022-04-14 12:26:43 [INFO]	[EVAL] Class Acc: 
[0.     0.6217 0.6435 0.6767 0.799  0.7619 0.8421 0.7266 0.9952 0.7753
 0.7934 0.917  0.7802 0.9715 0.7451 0.7686 0.8394 0.8891]
2022-04-14 12:26:44 [INFO]	[EVAL] The model with the best validation mIoU (0.6354) was saved at iter 400.
2022-04-14 12:26:50 [INFO]	[TRAIN] epoch: 8, iter: 410/1000, loss: 0.7373, lr: 0.006229, batch_cost: 0.5961, reader_cost: 0.00151, ips: 53.6830 samples/sec | ETA 00:05:51
2022-04-14 12:26:56 [INFO]	[TRAIN] epoch: 9, iter: 420/1000, loss: 0.6996, lr: 0.006134, batch_cost: 0.6003, reader_cost: 0.01966, ips: 53.3080 samples/sec | ETA 00:05:48
2022-04-14 12:27:02 [INFO]	[TRAIN] epoch: 9, iter: 430/1000, loss: 0.7439, lr: 0.006039, batch_cost: 0.6062, reader_cost: 0.00026, ips: 52.7906 samples/sec | ETA 00:05:45
2022-04-14 12:27:08 [INFO]	[TRAIN] epoch: 9, iter: 440/1000, loss: 0.8381, lr: 0.005944, batch_cost: 0.5920, reader_cost: 0.00020, ips: 54.0508 samples/sec | ETA 00:05:31
2022-04-14 12:27:14 [INFO]	[TRAIN] epoch: 9, iter: 450/1000, loss: 0.7520, lr: 0.005848, batch_cost: 0.5970, reader_cost: 0.00065, ips: 53.6005 samples/sec | ETA 00:05:28
2022-04-14 12:27:20 [INFO]	[TRAIN] epoch: 9, iter: 460/1000, loss: 0.8414, lr: 0.005753, batch_cost: 0.6021, reader_cost: 0.00129, ips: 53.1469 samples/sec | ETA 00:05:25
2022-04-14 12:27:26 [INFO]	[TRAIN] epoch: 10, iter: 470/1000, loss: 0.8407, lr: 0.005657, batch_cost: 0.5878, reader_cost: 0.01533, ips: 54.4415 samples/sec | ETA 00:05:11
2022-04-14 12:27:32 [INFO]	[TRAIN] epoch: 10, iter: 480/1000, loss: 0.7766, lr: 0.005561, batch_cost: 0.5939, reader_cost: 0.00088, ips: 53.8809 samples/sec | ETA 00:05:08
2022-04-14 12:27:38 [INFO]	[TRAIN] epoch: 10, iter: 490/1000, loss: 0.6825, lr: 0.005465, batch_cost: 0.6033, reader_cost: 0.00082, ips: 53.0438 samples/sec | ETA 00:05:07
2022-04-14 12:27:44 [INFO]	[TRAIN] epoch: 10, iter: 500/1000, loss: 0.6993, lr: 0.005369, batch_cost: 0.5977, reader_cost: 0.00220, ips: 53.5385 samples/sec | ETA 00:04:58
2022-04-14 12:27:44 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 143ms/step - batch_cost: 0.1428 - reader cost: 4.3921e-
2022-04-14 12:28:44 [INFO]	[EVAL] #Images: 420 mIoU: 0.6433 Acc: 0.8158 Kappa: 0.7854 
2022-04-14 12:28:44 [INFO]	[EVAL] Class IoU: 
[0.     0.5429 0.4983 0.6943 0.6581 0.5558 0.8024 0.4723 0.9608 0.6045
 0.5682 0.7844 0.6218 0.9213 0.6765 0.666  0.7303 0.8215]
2022-04-14 12:28:44 [INFO]	[EVAL] Class Acc: 
[0.     0.5851 0.7199 0.7802 0.8013 0.7535 0.8533 0.7028 0.9982 0.7434
 0.768  0.898  0.7861 0.9472 0.7861 0.7684 0.8051 0.9003]
2022-04-14 12:28:45 [INFO]	[EVAL] The model with the best validation mIoU (0.6433) was saved at iter 500.
2022-04-14 12:28:51 [INFO]	[TRAIN] epoch: 10, iter: 510/1000, loss: 0.7007, lr: 0.005272, batch_cost: 0.6111, reader_cost: 0.00197, ips: 52.3652 samples/sec | ETA 00:04:59
2022-04-14 12:28:56 [INFO]	[TRAIN] epoch: 10, iter: 520/1000, loss: 0.6916, lr: 0.005175, batch_cost: 0.5682, reader_cost: 0.00144, ips: 56.3156 samples/sec | ETA 00:04:32
2022-04-14 12:29:03 [INFO]	[TRAIN] epoch: 11, iter: 530/1000, loss: 0.6765, lr: 0.005078, batch_cost: 0.6610, reader_cost: 0.01684, ips: 48.4102 samples/sec | ETA 00:05:10
2022-04-14 12:29:09 [INFO]	[TRAIN] epoch: 11, iter: 540/1000, loss: 0.7006, lr: 0.004981, batch_cost: 0.6011, reader_cost: 0.00060, ips: 53.2397 samples/sec | ETA 00:04:36
2022-04-14 12:29:15 [INFO]	[TRAIN] epoch: 11, iter: 550/1000, loss: 0.6693, lr: 0.004884, batch_cost: 0.5991, reader_cost: 0.00319, ips: 53.4166 samples/sec | ETA 00:04:29
2022-04-14 12:29:21 [INFO]	[TRAIN] epoch: 11, iter: 560/1000, loss: 0.6371, lr: 0.004786, batch_cost: 0.6081, reader_cost: 0.00075, ips: 52.6256 samples/sec | ETA 00:04:27
2022-04-14 12:29:27 [INFO]	[TRAIN] epoch: 11, iter: 570/1000, loss: 0.6714, lr: 0.004688, batch_cost: 0.5879, reader_cost: 0.00069, ips: 54.4316 samples/sec | ETA 00:04:12
2022-04-14 12:29:33 [INFO]	[TRAIN] epoch: 12, iter: 580/1000, loss: 0.6552, lr: 0.004590, batch_cost: 0.5979, reader_cost: 0.01576, ips: 53.5168 samples/sec | ETA 00:04:11
2022-04-14 12:29:39 [INFO]	[TRAIN] epoch: 12, iter: 590/1000, loss: 0.7032, lr: 0.004492, batch_cost: 0.5926, reader_cost: 0.00015, ips: 53.9954 samples/sec | ETA 00:04:02
2022-04-14 12:29:45 [INFO]	[TRAIN] epoch: 12, iter: 600/1000, loss: 0.6684, lr: 0.004394, batch_cost: 0.5981, reader_cost: 0.00088, ips: 53.5068 samples/sec | ETA 00:03:59
2022-04-14 12:29:45 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 142ms/step - batch_cost: 0.1415 - reader cost: 4.0659e-0
2022-04-14 12:30:44 [INFO]	[EVAL] #Images: 420 mIoU: 0.6476 Acc: 0.8181 Kappa: 0.7884 
2022-04-14 12:30:44 [INFO]	[EVAL] Class IoU: 
[0.     0.5753 0.5149 0.6874 0.6624 0.5632 0.8085 0.4147 0.9513 0.6158
 0.6408 0.7854 0.6181 0.9192 0.6675 0.6941 0.7279 0.8105]
2022-04-14 12:30:44 [INFO]	[EVAL] Class Acc: 
[0.     0.7804 0.685  0.7804 0.806  0.7368 0.8634 0.8568 0.989  0.7441
 0.7843 0.9157 0.8449 0.9835 0.7029 0.7389 0.8188 0.874 ]
2022-04-14 12:30:45 [INFO]	[EVAL] The model with the best validation mIoU (0.6476) was saved at iter 600.
2022-04-14 12:30:51 [INFO]	[TRAIN] epoch: 12, iter: 610/1000, loss: 0.6927, lr: 0.004295, batch_cost: 0.6133, reader_cost: 0.00106, ips: 52.1739 samples/sec | ETA 00:03:59
2022-04-14 12:30:57 [INFO]	[TRAIN] epoch: 12, iter: 620/1000, loss: 0.6416, lr: 0.004196, batch_cost: 0.5878, reader_cost: 0.00156, ips: 54.4372 samples/sec | ETA 00:03:43
2022-04-14 12:31:03 [INFO]	[TRAIN] epoch: 13, iter: 630/1000, loss: 0.6696, lr: 0.004097, batch_cost: 0.5937, reader_cost: 0.01762, ips: 53.9003 samples/sec | ETA 00:03:39
2022-04-14 12:31:10 [INFO]	[TRAIN] epoch: 13, iter: 640/1000, loss: 0.6738, lr: 0.003997, batch_cost: 0.6349, reader_cost: 0.00130, ips: 50.4020 samples/sec | ETA 00:03:48
2022-04-14 12:31:16 [INFO]	[TRAIN] epoch: 13, iter: 650/1000, loss: 0.6204, lr: 0.003897, batch_cost: 0.5997, reader_cost: 0.00070, ips: 53.3608 samples/sec | ETA 00:03:29
2022-04-14 12:31:22 [INFO]	[TRAIN] epoch: 13, iter: 660/1000, loss: 0.5943, lr: 0.003797, batch_cost: 0.5970, reader_cost: 0.00115, ips: 53.6006 samples/sec | ETA 00:03:22
2022-04-14 12:31:28 [INFO]	[TRAIN] epoch: 13, iter: 670/1000, loss: 0.6692, lr: 0.003697, batch_cost: 0.5999, reader_cost: 0.00095, ips: 53.3394 samples/sec | ETA 00:03:17
2022-04-14 12:31:33 [INFO]	[TRAIN] epoch: 14, iter: 680/1000, loss: 0.6299, lr: 0.003596, batch_cost: 0.5912, reader_cost: 0.01579, ips: 54.1238 samples/sec | ETA 00:03:09
2022-04-14 12:31:40 [INFO]	[TRAIN] epoch: 14, iter: 690/1000, loss: 0.6693, lr: 0.003495, batch_cost: 0.6100, reader_cost: 0.00082, ips: 52.4570 samples/sec | ETA 00:03:09
2022-04-14 12:31:46 [INFO]	[TRAIN] epoch: 14, iter: 700/1000, loss: 0.6149, lr: 0.003394, batch_cost: 0.5998, reader_cost: 0.00157, ips: 53.3538 samples/sec | ETA 00:02:59
2022-04-14 12:31:46 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 143ms/step - batch_cost: 0.1428 - reader cost: 4.3040e-0
2022-04-14 12:32:46 [INFO]	[EVAL] #Images: 420 mIoU: 0.6746 Acc: 0.8354 Kappa: 0.8078 
2022-04-14 12:32:46 [INFO]	[EVAL] Class IoU: 
[0.     0.5989 0.5554 0.7251 0.6605 0.575  0.8147 0.4845 0.96   0.631
 0.6657 0.799  0.7254 0.9374 0.6921 0.7311 0.7419 0.8457]
2022-04-14 12:32:46 [INFO]	[EVAL] Class Acc: 
[0.     0.807  0.7245 0.8534 0.7856 0.7607 0.8763 0.7907 0.9987 0.7856
 0.7755 0.8704 0.8246 0.9818 0.7645 0.7729 0.8567 0.9267]
2022-04-14 12:32:46 [INFO]	[EVAL] The model with the best validation mIoU (0.6746) was saved at iter 700.
2022-04-14 12:32:52 [INFO]	[TRAIN] epoch: 14, iter: 710/1000, loss: 0.7421, lr: 0.003292, batch_cost: 0.5968, reader_cost: 0.00117, ips: 53.6205 samples/sec | ETA 00:02:53
2022-04-14 12:32:58 [INFO]	[TRAIN] epoch: 14, iter: 720/1000, loss: 0.6546, lr: 0.003190, batch_cost: 0.5940, reader_cost: 0.00032, ips: 53.8741 samples/sec | ETA 00:02:46
2022-04-14 12:33:04 [INFO]	[TRAIN] epoch: 15, iter: 730/1000, loss: 0.5932, lr: 0.003088, batch_cost: 0.5865, reader_cost: 0.01657, ips: 54.5621 samples/sec | ETA 00:02:38
2022-04-14 12:33:10 [INFO]	[TRAIN] epoch: 15, iter: 740/1000, loss: 0.5941, lr: 0.002985, batch_cost: 0.5966, reader_cost: 0.00024, ips: 53.6345 samples/sec | ETA 00:02:35
2022-04-14 12:33:16 [INFO]	[TRAIN] epoch: 15, iter: 750/1000, loss: 0.6004, lr: 0.002882, batch_cost: 0.6279, reader_cost: 0.00230, ips: 50.9615 samples/sec | ETA 00:02:36
2022-04-14 12:33:23 [INFO]	[TRAIN] epoch: 15, iter: 760/1000, loss: 0.6031, lr: 0.002779, batch_cost: 0.6034, reader_cost: 0.00289, ips: 53.0347 samples/sec | ETA 00:02:24
2022-04-14 12:33:28 [INFO]	[TRAIN] epoch: 15, iter: 770/1000, loss: 0.6118, lr: 0.002675, batch_cost: 0.5960, reader_cost: 0.00167, ips: 53.6929 samples/sec | ETA 00:02:17
2022-04-14 12:33:34 [INFO]	[TRAIN] epoch: 15, iter: 780/1000, loss: 0.6142, lr: 0.002570, batch_cost: 0.5754, reader_cost: 0.00081, ips: 55.6109 samples/sec | ETA 00:02:06
2022-04-14 12:33:40 [INFO]	[TRAIN] epoch: 16, iter: 790/1000, loss: 0.5906, lr: 0.002465, batch_cost: 0.6159, reader_cost: 0.01718, ips: 51.9571 samples/sec | ETA 00:02:09
2022-04-14 12:33:46 [INFO]	[TRAIN] epoch: 16, iter: 800/1000, loss: 0.5972, lr: 0.002360, batch_cost: 0.6070, reader_cost: 0.00100, ips: 52.7145 samples/sec | ETA 00:02:01
2022-04-14 12:33:46 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 143ms/step - batch_cost: 0.1431 - reader cost: 4.3473e-0
2022-04-14 12:34:47 [INFO]	[EVAL] #Images: 420 mIoU: 0.6764 Acc: 0.8370 Kappa: 0.8102 
2022-04-14 12:34:47 [INFO]	[EVAL] Class IoU: 
[0.     0.6189 0.56   0.7151 0.6693 0.5738 0.7879 0.5127 0.9634 0.6378
 0.669  0.8069 0.7086 0.9449 0.6956 0.7226 0.7423 0.8469]
2022-04-14 12:34:47 [INFO]	[EVAL] Class Acc: 
[0.     0.7452 0.7343 0.7984 0.8166 0.767  0.8257 0.7523 1.     0.7964
 0.7517 0.8972 0.8457 0.9782 0.7789 0.7646 0.8452 0.9181]
2022-04-14 12:34:48 [INFO]	[EVAL] The model with the best validation mIoU (0.6764) was saved at iter 800.
2022-04-14 12:34:54 [INFO]	[TRAIN] epoch: 16, iter: 810/1000, loss: 0.5991, lr: 0.002254, batch_cost: 0.5941, reader_cost: 0.00210, ips: 53.8640 samples/sec | ETA 00:01:52
2022-04-14 12:34:59 [INFO]	[TRAIN] epoch: 16, iter: 820/1000, loss: 0.5848, lr: 0.002147, batch_cost: 0.5925, reader_cost: 0.00075, ips: 54.0113 samples/sec | ETA 00:01:46
2022-04-14 12:35:05 [INFO]	[TRAIN] epoch: 16, iter: 830/1000, loss: 0.5802, lr: 0.002040, batch_cost: 0.5780, reader_cost: 0.00066, ips: 55.3610 samples/sec | ETA 00:01:38
2022-04-14 12:35:11 [INFO]	[TRAIN] epoch: 17, iter: 840/1000, loss: 0.6143, lr: 0.001933, batch_cost: 0.6128, reader_cost: 0.01615, ips: 52.2204 samples/sec | ETA 00:01:38
2022-04-14 12:35:17 [INFO]	[TRAIN] epoch: 17, iter: 850/1000, loss: 0.6197, lr: 0.001824, batch_cost: 0.6147, reader_cost: 0.00096, ips: 52.0584 samples/sec | ETA 00:01:32
2022-04-14 12:35:24 [INFO]	[TRAIN] epoch: 17, iter: 860/1000, loss: 0.5561, lr: 0.001715, batch_cost: 0.6057, reader_cost: 0.00072, ips: 52.8288 samples/sec | ETA 00:01:24
2022-04-14 12:35:30 [INFO]	[TRAIN] epoch: 17, iter: 870/1000, loss: 0.5752, lr: 0.001605, batch_cost: 0.5980, reader_cost: 0.00125, ips: 53.5096 samples/sec | ETA 00:01:17
2022-04-14 12:35:35 [INFO]	[TRAIN] epoch: 17, iter: 880/1000, loss: 0.5779, lr: 0.001495, batch_cost: 0.5963, reader_cost: 0.00278, ips: 53.6679 samples/sec | ETA 00:01:11
2022-04-14 12:35:41 [INFO]	[TRAIN] epoch: 18, iter: 890/1000, loss: 0.5226, lr: 0.001383, batch_cost: 0.5988, reader_cost: 0.01577, ips: 53.4440 samples/sec | ETA 00:01:05
2022-04-14 12:35:47 [INFO]	[TRAIN] epoch: 18, iter: 900/1000, loss: 0.5943, lr: 0.001270, batch_cost: 0.5943, reader_cost: 0.00091, ips: 53.8452 samples/sec | ETA 00:00:59
2022-04-14 12:35:47 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 143ms/step - batch_cost: 0.1424 - reader cost: 4.8283e-
2022-04-14 12:36:47 [INFO]	[EVAL] #Images: 420 mIoU: 0.6823 Acc: 0.8382 Kappa: 0.8114 
2022-04-14 12:36:47 [INFO]	[EVAL] Class IoU: 
[0.     0.6235 0.5526 0.7267 0.6715 0.5952 0.8155 0.516  0.964  0.6368
 0.6684 0.8126 0.7114 0.9239 0.699  0.7702 0.7441 0.8504]
2022-04-14 12:36:47 [INFO]	[EVAL] Class Acc: 
[0.     0.7182 0.683  0.8378 0.8101 0.7393 0.8766 0.763  0.9997 0.8103
 0.7642 0.8939 0.8346 0.9864 0.7791 0.8363 0.8568 0.9297]
2022-04-14 12:36:48 [INFO]	[EVAL] The model with the best validation mIoU (0.6823) was saved at iter 900.
2022-04-14 12:36:54 [INFO]	[TRAIN] epoch: 18, iter: 910/1000, loss: 0.5703, lr: 0.001156, batch_cost: 0.6115, reader_cost: 0.00312, ips: 52.3318 samples/sec | ETA 00:00:55
2022-04-14 12:37:00 [INFO]	[TRAIN] epoch: 18, iter: 920/1000, loss: 0.6093, lr: 0.001041, batch_cost: 0.6014, reader_cost: 0.00108, ips: 53.2099 samples/sec | ETA 00:00:48
2022-04-14 12:37:06 [INFO]	[TRAIN] epoch: 18, iter: 930/1000, loss: 0.5604, lr: 0.000925, batch_cost: 0.5959, reader_cost: 0.00160, ips: 53.7019 samples/sec | ETA 00:00:41
2022-04-14 12:37:12 [INFO]	[TRAIN] epoch: 19, iter: 940/1000, loss: 0.5809, lr: 0.000807, batch_cost: 0.5947, reader_cost: 0.01608, ips: 53.8086 samples/sec | ETA 00:00:35
2022-04-14 12:37:18 [INFO]	[TRAIN] epoch: 19, iter: 950/1000, loss: 0.5609, lr: 0.000687, batch_cost: 0.5905, reader_cost: 0.00034, ips: 54.1959 samples/sec | ETA 00:00:29
2022-04-14 12:37:24 [INFO]	[TRAIN] epoch: 19, iter: 960/1000, loss: 0.6061, lr: 0.000564, batch_cost: 0.6175, reader_cost: 0.00174, ips: 51.8181 samples/sec | ETA 00:00:24
2022-04-14 12:37:30 [INFO]	[TRAIN] epoch: 19, iter: 970/1000, loss: 0.5418, lr: 0.000439, batch_cost: 0.6037, reader_cost: 0.00178, ips: 53.0053 samples/sec | ETA 00:00:18
2022-04-14 12:37:36 [INFO]	[TRAIN] epoch: 19, iter: 980/1000, loss: 0.5613, lr: 0.000309, batch_cost: 0.5929, reader_cost: 0.00152, ips: 53.9685 samples/sec | ETA 00:00:11
2022-04-14 12:37:42 [INFO]	[TRAIN] epoch: 20, iter: 990/1000, loss: 0.5241, lr: 0.000173, batch_cost: 0.5946, reader_cost: 0.01715, ips: 53.8205 samples/sec | ETA 00:00:05
2022-04-14 12:37:48 [INFO]	[TRAIN] epoch: 20, iter: 1000/1000, loss: 0.5681, lr: 0.000020, batch_cost: 0.5928, reader_cost: 0.00015, ips: 53.9844 samples/sec | ETA 00:00:00
2022-04-14 12:37:48 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
420/420 [==============================] - 60s 144ms/step - batch_cost: 0.1434 - reader cost: 5.1340e-
2022-04-14 12:38:48 [INFO]	[EVAL] #Images: 420 mIoU: 0.6804 Acc: 0.8398 Kappa: 0.8132 
2022-04-14 12:38:48 [INFO]	[EVAL] Class IoU: 
[0.     0.6279 0.5664 0.7303 0.6736 0.5809 0.8024 0.5217 0.963  0.6403
 0.6745 0.8123 0.6961 0.921  0.699  0.7392 0.7469 0.8508]
2022-04-14 12:38:48 [INFO]	[EVAL] Class Acc: 
[0.     0.7598 0.7189 0.842  0.8211 0.775  0.8507 0.7527 1.     0.782
 0.7552 0.8984 0.8332 0.9858 0.7765 0.804  0.8528 0.936 ]
2022-04-14 12:38:49 [INFO]	[EVAL] The model with the best validation mIoU (0.6823) was saved at iter 900.
<class 'paddle.nn.layer.conv.Conv2D'>'s flops has been counted
Customize Function has been applied to <class 'paddle.nn.layer.norm.SyncBatchNorm'>
Cannot find suitable count function for <class 'paddleseg.models.ocrnet.SpatialGatherBlock'>. Treat it as zero FLOPs.
Cannot find suitable count function for <class 'paddle.nn.layer.common.Dropout2D'>. Treat it as zero FLOPs.
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/tensor/creation.py:125: DeprecationWarning: `np.object` is a deprecated alias for the builtin `object`. To silence this warning, use `object` by itself. Doing this will not modify any behavior and is safe. 
Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
  if data.dtype == np.object:
Total Flops: 1757012736     Total Params: 12122446
登录后复制
       

模型测试

参考paddleSeg

In [9]
# 模型测试, aug_eval可以开启增强测试模式# !python val.py \#        --config configs/unet/unet_dlrsd.yml \#        --model_path output/best_model/model.pdparams \#        --aug_eval \#        --scales 0.75 1.0 1.25 \#        --flip_horizontal# 使用ocrnet进行测试!python val.py \
       --config ocrnet_dlrsd.yml \
       --model_path output/best_model/model.pdparams \
       --aug_eval \
       --scales 0.75 1.0 1.25 \
       --flip_horizontal
登录后复制
       
2022-04-14 12:38:53 [INFO]
---------------Config Information---------------
batch_size: 2
iters: 160000
loss:
  coef:
  - 1
  - 0.4
  types:
  - type: CrossEntropyLoss
  - type: CrossEntropyLoss
lr_scheduler:
  end_lr: 0
  learning_rate: 0.01
  power: 0.9
  type: PolynomialDecay
model:
  backbone:
    pretrained: https://bj.bcebos.com/paddleseg/dygraph/hrnet_w18_ssld.tar.gz
    type: HRNet_W18
  backbone_indices:
  - 0
  num_classes: 18
  type: OCRNet
optimizer:
  momentum: 0.9
  type: sgd
  weight_decay: 4.0e-05
train_dataset:
  dataset_root: dataset/dlrsd
  mode: train
  num_classes: 18
  train_path: dataset/dlrsd/train.txt
  transforms:
  - max_scale_factor: 2.0
    min_scale_factor: 0.5
    scale_step_size: 0.25
    type: ResizeStepScaling
  - crop_size:
    - 256
    - 256
    type: RandomPaddingCrop
  - type: RandomHorizontalFlip
  - type: Normalize
  type: Dataset
val_dataset:
  dataset_root: dataset/dlrsd
  mode: val
  num_classes: 18
  transforms:
  - type: Normalize
  - target_size:
    - 256
    - 256
    type: Resize
  type: Dataset
  val_path: dataset/dlrsd/val.txt
------------------------------------------------
W0414 12:38:53.728370  4958 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.0, Runtime API Version: 10.1
W0414 12:38:53.728415  4958 device_context.cc:422] device: 0, cuDNN Version: 7.6.
2022-04-14 12:38:57 [INFO]	Loading pretrained model from https://bj.bcebos.com/paddleseg/dygraph/hrnet_w18_ssld.tar.gz
2022-04-14 12:38:58 [INFO]	There are 1525/1525 variables loaded into HRNet.
2022-04-14 12:38:58 [INFO]	Loading pretrained model from output/best_model/model.pdparams
2022-04-14 12:38:58 [INFO]	There are 1583/1583 variables loaded into OCRNet.
2022-04-14 12:38:58 [INFO]	Loaded trained params of model successfully
2022-04-14 12:38:58 [INFO]	Start evaluating (total_samples: 420, total_iters: 420)...
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/dygraph/math_op_patch.py:239: UserWarning: The dtype of left and right variables are not the same, left dtype is paddle.int32, but right dtype is paddle.bool, the right dtype will convert to paddle.int32
  format(lhs_dtype, rhs_dtype, lhs_dtype))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/dygraph/math_op_patch.py:239: UserWarning: The dtype of left and right variables are not the same, left dtype is paddle.int64, but right dtype is paddle.bool, the right dtype will convert to paddle.int64
  format(lhs_dtype, rhs_dtype, lhs_dtype))
420/420 [==============================] - 335s 797ms/step - batch_cost: 0.7971 - reader cost: 8.0393e-
2022-04-14 12:44:33 [INFO]	[EVAL] #Images: 420 mIoU: 0.6839 Acc: 0.8405 Kappa: 0.8143 
2022-04-14 12:44:33 [INFO]	[EVAL] Class IoU: 
[0.     0.6325 0.563  0.7203 0.6722 0.6052 0.8156 0.5139 0.9655 0.6445
 0.6614 0.814  0.7313 0.9119 0.7007 0.7548 0.7473 0.8563]
2022-04-14 12:44:33 [INFO]	[EVAL] Class Acc: 
[0.     0.7304 0.7056 0.8246 0.7997 0.7365 0.872  0.8375 1.     0.8088
 0.7385 0.9016 0.829  0.9841 0.7464 0.8043 0.8618 0.9425]
登录后复制
       

模型推理

参考paddleSeg

In [10]
# 模型推理, image_path可以设置为单个文件或者文件夹# !python predict.py \#        --config configs/unet/unet_dlrsd.yml \#        --model_path output/best_model/model.pdparams \#        --image_path dataset/dlrsd/image/agricultural00.tif \#        --save_dir output/result# 使用ocrnet模型推理!python predict.py \
       --config ocrnet_dlrsd.yml \
       --model_path output/best_model/model.pdparams \
       --image_path dataset/dlrsd/image/agricultural00.tif \
       --save_dir output/result
登录后复制
       
2022-04-14 12:44:36 [INFO]
---------------Config Information---------------
batch_size: 2
iters: 160000
loss:
  coef:
  - 1
  - 0.4
  types:
  - type: CrossEntropyLoss
  - type: CrossEntropyLoss
lr_scheduler:
  end_lr: 0
  learning_rate: 0.01
  power: 0.9
  type: PolynomialDecay
model:
  backbone:
    pretrained: https://bj.bcebos.com/paddleseg/dygraph/hrnet_w18_ssld.tar.gz
    type: HRNet_W18
  backbone_indices:
  - 0
  num_classes: 18
  type: OCRNet
optimizer:
  momentum: 0.9
  type: sgd
  weight_decay: 4.0e-05
train_dataset:
  dataset_root: dataset/dlrsd
  mode: train
  num_classes: 18
  train_path: dataset/dlrsd/train.txt
  transforms:
  - max_scale_factor: 2.0
    min_scale_factor: 0.5
    scale_step_size: 0.25
    type: ResizeStepScaling
  - crop_size:
    - 256
    - 256
    type: RandomPaddingCrop
  - type: RandomHorizontalFlip
  - type: Normalize
  type: Dataset
val_dataset:
  dataset_root: dataset/dlrsd
  mode: val
  num_classes: 18
  transforms:
  - type: Normalize
  - target_size:
    - 256
    - 256
    type: Resize
  type: Dataset
  val_path: dataset/dlrsd/val.txt
------------------------------------------------
W0414 12:44:36.429344  6170 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.0, Runtime API Version: 10.1
W0414 12:44:36.429426  6170 device_context.cc:422] device: 0, cuDNN Version: 7.6.
2022-04-14 12:44:39 [INFO]	Loading pretrained model from https://bj.bcebos.com/paddleseg/dygraph/hrnet_w18_ssld.tar.gz
2022-04-14 12:44:40 [INFO]	There are 1525/1525 variables loaded into HRNet.
2022-04-14 12:44:40 [INFO]	Number of predict images = 1
2022-04-14 12:44:40 [INFO]	Loading pretrained model from output/best_model/model.pdparams
2022-04-14 12:44:41 [INFO]	There are 1583/1583 variables loaded into OCRNet.
2022-04-14 12:44:41 [INFO]	Start to predict...
1/1 [==============================] - 0s 173ms/step
登录后复制
       

推理结果展示

基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例 - php中文网        

基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例 - php中文网        

基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例 - php中文网        

以上就是基于paddleSeg的自定义遥感数据语义分割——以DLRSD为例的详细内容，更多请关注php中文网其它相关文章！

相关标签：
linux python git ai cos yy red igs
最佳 Windows 性能的顶级免费优化软件
最佳 Windows 性能的顶级免费优化软件

每个人都需要一台速度更快、更稳定的 PC。随着时间的推移，垃圾文件、旧注册表数据和不必要的后台进程会占用资源并降低性能。幸运的是，许多工具可以让 Windows 保持平稳运行。

下载
来源：php中文网
收藏 点赞
上一篇：趣味体验馆踩坑指南 下一篇：【AI达人创造营第二期】电机异音AI诊断
本文内容由网友自发贡献，版权归原作者所有，本站不承担相应法律责任。如您发现有涉嫌抄袭侵权的内容，请联系admin@php.cn
作者最新文章
最新问题
2024免费AI做视频入口 最好用的AI视频生成器 2024免费AI做视频入口是https://pixverse.ai/，该平台新用户注册送100积分且每日登录领50积分，采用Diffusion+Transformer技术，支持文生视频、图生视频及多片段生成，可输出5秒或8秒高清短视频，满足社交平台需求。
2025-11-13 19:57:36
343
文心一言网页版登录地址 文心一言账号注册入口 文心一言网页版登录地址为https://yiyan.baidu.com，注册需通过百度账号完成，使用手机号获取验证码设置密码即可，支持扫码、账号密码或短信登录，具备多轮对话、内容创作、图文处理及智能决策辅助功能。
2025-11-13 19:52:03
532
ai志愿助手官方网站访问入口_ai志愿助手平台直达链接官方首页 答案：主流AI志愿助手入口包括百度、支付宝、夸克和掌上高考。通过百度APP搜“AI志愿助手”、支付宝搜“高考”进入小程序、夸克APP使用“夸克高考”频道，或访问掌上高考官网gkcx.eol.cn获取权威服务，建议结合多平台数据并核对官方信息以优化志愿填报。
2025-11-13 19:51:07
184
百度AI搜索入门指南 附官方免登录访问入口 百度AI搜索提供深度搜索与多模态交互，支持智能/普通模式切换、逻辑推理及AI生图修图，依托文心大模型实现文案生成与图像处理，用户可免登录通过官网直接访问，界面简洁适配多设备。
2025-11-13 19:45:23
653
哩布哩布AI怎么使用局部重绘_哩布哩布AI蒙版修改指定区域教程 哩布哩布AI的局部重绘功能可通过蒙版精准修改图像区域。1、上传图像后点击“局部重绘”，使用画笔绘制需修改区域，调整画笔大小与不透明度；2、设置重绘强度（建议0.4~0.7）、输入新描述提示词，并选择是否保留颜色信息；3、预览蒙版与融合效果，确认后生成高分辨率图像；4、通过羽化或二次编辑优化边缘与细节；5、最终导出为PNG或JPG格式，适用于修复、换装、调脸等场景。
2025-11-13 19:43:18
834
30 亿美元打造现实版“星门”！OpenAI 绝密AI基地曝光 在人工智能领域的算力争夺战中，OpenAI再次掀起波澜。近期，一笔高达30亿美元的专项债务融资正式注入其代号为Stargate（星门）的下一代AI基础设施计划。这笔资金由华尔街顶尖另类资产管理公司BlueOwl提供，将全部用于在美国新墨西哥州建设一座前所未有的超大规模人工智能数据中心。该中心的核心任务是为GPT系列大模型、SANA视频生成系统以及未来更先进的AI智能体提供强大而稳定的算力支撑。自规划之初，Stargate项目就被视为OpenAI破解算力瓶颈的战略重器。这并非传统意义上的数据中心，
2025-11-13 19:42:12
266
天宫AI如何登录_天宫AI登录入口与账号绑定方法 首先确认访问天宫AI官网https://tiangong.cn，通过扫码、账号密码或第三方方式登录，若忘记密码可找回，同时可在账号安全设置中绑定或更换手机号与邮箱。
2025-11-13 19:41:43
390
讯飞火星AI怎样实现中英互译_讯飞火星AI中英翻译功能使用与效果提升方法 启用实时对话翻译、使用文本输入、自定义术语库、调整语音灵敏度、开启上下文记忆可提升讯飞火星AI中英互译准确性与响应速度。
2025-11-13 19:39:07
685
官方公布小米汽车最新奖项：获得2025最佳结构奖 11月13日，小米汽车正式宣布其YU7车型所搭载的20合1一体化压铸铝三角梁，在北美压铸协会（NADCA）主办的2025InternationalDieCastingCompetition（2025国际压铸大赛）中斩获“最佳结构奖”，该奖项为本次赛事的最高荣誉。小米YU7据官方介绍，这款一体化压铸铝三角梁是车辆碰撞力传递路径中的核心组件。在发生碰撞时，能够将冲击力有效传导至车身对侧，显著增强整车吸能能力，减少对乘员舱的直接冲击。此外，该部件创新性地融合了车身结构与热管理系统的跨域集成设计，集成了
2025-11-13 19:31:16
420
欧拉5正式开启预售 10.98万起 推五款车型主打年轻市场 11月12日，长城汽车旗下欧拉品牌正式启动其首款纯电动SUV——欧拉5的预售，共推出五个配置车型，预售价覆盖10.98万元至14.28万元区间。作为专为年轻消费群体打造的“严选级”产品，欧拉5在外观设计、智能化配置以及续航表现上均展现出不俗实力。此次发布的车型涵盖480kmAir版（10.98万元）、480kmPro版（11.98万元）、580kmMax版（12.98万元）、480km激光雷达版（13.28万元）和580km激光雷达版（14.28万元）。全系搭载由蜂巢能源供应的磷酸铁锂电池组，提
2025-11-13 19:23:32
674
相关专题
更多>
热门推荐
开源免费商场系统广告
热门教程
更多>
相关推荐
热门推荐
最新课程
最新下载
更多>
网站特效
网站源码
网站素材
前端模板
关于我们 免责申明 举报中心 意见反馈 讲师合作 广告合作 最新更新 English
php中文网：公益在线php培训，帮助PHP学习者快速成长！
关注服务号 技术交流群
PHP中文网订阅号
每天精选资源文章推送
PHP中文网APP
随时随地碎片化学习

Copyright 2014-2025 https://www.php.cn/ All Rights Reserved | php.cn | 湘ICP备2023035733号

  • PHP学习

  • 技术支持

  • 返回顶部