Suck the cat with code! This paper is participating in[Cat Essay Campaign]

I. Paddleseg-based cat master and child image segmentation

1. PaddleSeg profile

PaddleSeg is an end-to-end image segmentation development kit based on PaddlePaddle, which covers a large number of high quality segmentation models in different directions, including high precision and lightweight. Through the modular design, it provides two application methods of configuration-driven and API call, which helps developers to complete the whole process of image segmentation application from training to deployment more conveniently.

features

High-precision model: The high-precision backbone network is obtained based on the semi-supervised label knowledge distillation scheme (SSLD) training developed by Baidu itself, and 50+ high-quality pre-training model is provided in combination with cutting-edge segmentation technology. The effect is better than other open source implementations.
Modular design: support 15+ mainstream segmentation network, combined with modular design of different components such as data enhancement strategy, backbone network, loss function, developers can assemble a variety of training configurations based on actual application scenarios, to meet different performance and accuracy requirements.
High performance: support multi-process asynchronous I/O, multi-card parallel training, evaluation and other acceleration strategies, combined with the video memory optimization function of the flying-blade core framework, can greatly reduce the training cost of segmentation model, so that developers can complete image segmentation training at a lower cost and more efficiently.

! git clone https://gitee.com/paddlepaddle/PaddleSeg.git  --depth=1
Copy the code

Cloning into 'PaddleSeg'... remote: Enumerating objects: 1589, done.[K remote: Counting objects: 100% (1589/1589), done.[K remote: Compressing objects: 100% (1354/1354), done.[K remote: Total 1589 (delta 309), reused 1117 (delta 142), pack-reused 0[K Receiving objects: 100% (1589/1589) and 88.49 MiB | 5.57 MiB/s, done. Resolving deltas: 100% (309/309), done. Checking connectivity... done.Copy the code

2. Data set making

Unzip the dataset! mkdir dataset ! tar -xvf data/data50154/images.tar.gz -C dataset/ ! tar -xvf data/data50154/annotations.tar.gz -C dataset/Copy the code

# delete the list. TXT header
import pandas as pd
import shutil
import os


# Image CLASS-ID SPECIES BREED ID
# ID: 1:37 Class ids
# SPECIES: 1:Cat 2:Dog
# BREED ID: 1-25:Cat 1:12:Dog
# All images with 1st letter as captial are cat images
# images with small first letter are dog images
# ._Abyssinian_100.png

def copyfile(animal, filename) :
    # image \ label list
    file_list = []
    image_file = filename + '.jpg'
    label_file = filename + '.png'

    if os.path.exists(os.path.join('dataset/images', image_file)):
        shutil.copy(os.path.join('dataset/images', image_file), os.path.join(f'{animal}/images', image_file))
        shutil.copy(os.path.join('dataset/annotations/trimaps', label_file),
                    os.path.join(f'{animal}/labels', label_file))
        temp = os.path.join('images/', image_file) + ' ' + os.path.join('labels/',label_file) + '\n'
        file_list.append(temp)
    with open(os.path.join(animal, animal + '.txt'), 'a') as f:
        f.writelines(file_list)


if __name__ == "__main__":

    data = pd.read_csv('dataset/annotations/list.txt', header=None, sep=' ')
    data.head()

    cat = data[data[2] = =1]
    dog = data[data[2] = =2]

    for item in cat[0]:
        copyfile('cat', item)

    for item in dog[0]:
        copyfile('dog', item)

Copy the code

Delete unnecessary data! rm dataset/ -rfCopy the code

3. Train custom data sets

3.1 File Structure

├ ─ ─ the TXT ├ ─ ─ images │ ├ ─ ─ Abyssinian_100. JPG │ ├ ─ ─ Abyssinian_101. JPG │ ├ ─ ─... ├ ─ ─ labels │ ├ ─ ─ Abyssinian_100. PNG │ ├ ─ ─ Abyssinian_101. PNG │ ├ ─ ─...Copy the code

3.2 Contents of cat.txt:

images/Abyssinian_1.jpg labels/Abyssinian_1.png
images/Abyssinian_10.jpg labels/Abyssinian_10.png
images/Abyssinian_100.jpg labels/Abyssinian_100.png
...
Copy the code

3. Data viewing

%cd ~
from PIL import Image

img=Image.open('cat/images/Abyssinian_101.jpg')
print(img)
img
Copy the code

/home/aistudio
<PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=450x313 at 0x7F157A098A90>
Copy the code

img=Image.open('cat/images/Abyssinian_123.jpg')
print(img)
img
Copy the code

<PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=500x333 at 0x7F174201AC90>
Copy the code

4. Label processing

The tags are sorted from 0. Data for this project was extracted from Oxford-IIIT Pet www.robots.ox.ac.uk/~vgg/data/p… Pet data set, which is encoded from 1, so it needs to be re-encoded. The background is set to 0 and the image to 1.

import pandas as pd
import os
import matplotlib.pyplot as plt
import numpy as np
from PIL import Image

def re_label(filename) :
    img = plt.imread(filename) * 255.0
    img_label = np.zeros((img.shape[0], img.shape[1]), np.uint8)
    for i in range(img.shape[0) :for j in range(img.shape[1]):
            value = img[i, j]
            if value == 2:
                img_label[i, j] = 1
    label0 = Image.fromarray(np.uint8(img_label))
    label0.save( filename)

data=pd.read_csv("cat/cat.txt", header=None, sep=' ') 
for item in data[1]:
    re_label(os.path.join('cat', item))
print('Done! ')    
Copy the code

Done!Copy the code

2. Data set preprocessing

import os
from sklearn.model_selection import train_test_split
import pandas as pd


def break_data(target, rate=0.2) :
    origin_dataset = pd.read_csv("cat/cat.txt", header=None, sep=' ')  # add parameter
    train_data, test_data = train_test_split(origin_dataset, test_size=rate)
    train_data,eval_data=train_test_split(train_data, test_size=rate)
    train_filename = os.path.join(target, 'train.txt')
    test_filename = os.path.join(target, 'test.txt')
    eval_filename = os.path.join(target, 'eval.txt')

    train_data.to_csv(train_filename, index=False, sep=' ',  header=None)
    test_data.to_csv(test_filename, index=False, sep=' ', header=None)
    eval_data.to_csv(eval_filename, index=False, sep=' ', header=None)

    print('train_data:'.len(train_data))
    print('test_data:'.len(test_data))
    print('eval_data:'.len(eval_data))

if __name__ == '__main__':
    break_data(target='cat', rate=0.2)
Copy the code

train_data: 1516
test_data: 475
eval_data: 380
Copy the code

# check! head cat/train.txtCopy the code

images/Bengal_173.jpg labels/Bengal_173.png
images/Siamese_179.jpg labels/Siamese_179.png
images/British_Shorthair_201.jpg labels/British_Shorthair_201.png
images/Russian_Blue_60.jpg labels/Russian_Blue_60.png
images/British_Shorthair_93.jpg labels/British_Shorthair_93.png
images/British_Shorthair_26.jpg labels/British_Shorthair_26.png
images/British_Shorthair_209.jpg labels/British_Shorthair_209.png
images/British_Shorthair_101.jpg labels/British_Shorthair_101.png
images/British_Shorthair_269.jpg labels/British_Shorthair_269.png
images/Ragdoll_59.jpg labels/Ragdoll_59.png
Copy the code

Three, configuration,

! cp PaddleSeg/configs/quick_start/bisenet_optic_disc_512x512_1k.yml ~/bisenet_optic_disc_512x512_1k.ymlCopy the code

To modify bisenet_optic_disc_512x512_1k.yml, note the following:

1. Set the path of the data set
2. Num_classes setting, background does not count
3. Transforms Settings
4. Loss Settings

batch_size: 256 iters: 500 train_dataset: type: Dataset dataset_root: /home/aistudio/cat/ train_path: /home/aistudio/cat/train.txt num_classes: 2 transforms: - type: ResizeStepScaling min_scale_factor: Max_scale_factor: 2.0 scale_step_size: 0.25 - type: RandomPaddingCrop crop_size: [224, 224] - type: Randomhorizontalflip-type: RandomDistort brightness_range: 0.4 Contrast_range: 0.4 saturation_range: 0.4 - type: Normalize mode: train val_dataset: type: Dataset dataset_root: /home/aistudio/cat/ val_path: /home/aistudio/cat/eval.txt num_classes: 2 transforms: - type: Normalize mode: val optimizer: type: sgd momentum: 0.9 weight_decay: 0.0005 lR_scheduler: type: PolynomialDecay Learning_rate: 0.05 end_LR: 0 power: 0.9 Loss: types: - type: CrossEntropyLoss coef: [1] model: type: FCN backbone: type: HRNet_W18_Small_V1 align_corners: False num_classes: 2 pretrained: NullCopy the code

Fourth, training

%cd ~/PaddleSeg/ ! python train.py --config .. /bisenet_optic_disc_512x512_1k.yml\ --do_eval \ --use_vdl \ --save_interval50 \
    --save_dir output
Copy the code

/home/aistudio/PaddleSeg
2021-11-03 00:21:10 [INFO]	
------------Environment Information-------------
platform: Linux-4.4.0-150-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-SXM2-32GB']
GCC: gcc (Ubuntu 7.5.0-3ubuntu1~16.04) 7.5.0
PaddlePaddle: 2.1.2
OpenCV: 4.1.1
------------------------------------------------
2021-11-03 00:21:10 [INFO]	
---------------Config Information---------------
batch_size: 256
iters: 500
loss:
  coef:
  - 1
  types:
  - ignore_index: 255
    type: CrossEntropyLoss
lr_scheduler:
  end_lr: 0
  learning_rate: 0.05
  power: 0.9
  type: PolynomialDecay
model:
  backbone:
    align_corners: false
    type: HRNet_W18_Small_V1
  num_classes: 2
  pretrained: null
  type: FCN
optimizer:
  momentum: 0.9
  type: sgd
  weight_decay: 0.0005
train_dataset:
  dataset_root: /home/aistudio/cat/
  mode: train
  num_classes: 2
  train_path: /home/aistudio/cat/train.txt
  transforms:
  - max_scale_factor: 2.0
    min_scale_factor: 0.5
    scale_step_size: 0.25
    type: ResizeStepScaling
  - crop_size:
    - 224
    - 224
    type: RandomPaddingCrop
  - type: RandomHorizontalFlip
  - brightness_range: 0.4
    contrast_range: 0.4
    saturation_range: 0.4
    type: RandomDistort
  - type: Normalize
  type: Dataset
val_dataset:
  dataset_root: /home/aistudio/cat/
  mode: val
  num_classes: 2
  transforms:
  - type: Normalize
  type: Dataset
  val_path: /home/aistudio/cat/eval.txt
------------------------------------------------
W1103 00:21:10.323276  1991 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 10.1, Runtime API Version: 10.1
W1103 00:21:10.323315  1991 device_context.cc:422] device: 0, cuDNN Version: 7.6.
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/nn/layer/norm.py:641: UserWarning: When training, we now always track global mean and variance.
  "When training, we now always track global mean and variance.")
/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.float32, but right dtype is paddle.int64, the right dtype will convert to paddle.float32
  format(lhs_dtype, rhs_dtype, lhs_dtype))
2021-11-03 00:22:00 [INFO]	[TRAIN] epoch: 2, iter: 10/500, loss: 0.6637, lr: 0.049189, batch_cost: 4.0507, reader_cost: 3.05866, ips: 63.1989 samples/sec | ETA 00:33:04
2021-11-03 00:22:39 [INFO]	[TRAIN] epoch: 4, iter: 20/500, loss: 0.6265, lr: 0.048287, batch_cost: 3.9186, reader_cost: 2.92979, ips: 65.3300 samples/sec | ETA 00:31:20
2021-11-03 00:23:19 [INFO]	[TRAIN] epoch: 6, iter: 30/500, loss: 0.6230, lr: 0.047382, batch_cost: 3.9892, reader_cost: 2.99757, ips: 64.1727 samples/sec | ETA 00:31:14
2021-11-03 00:24:01 [INFO]	[TRAIN] epoch: 8, iter: 40/500, loss: 0.6111, lr: 0.046476, batch_cost: 4.1665, reader_cost: 3.17214, ips: 61.4422 samples/sec | ETA 00:31:56
2021-11-03 00:24:41 [INFO]	[TRAIN] epoch: 10, iter: 50/500, loss: 0.6139, lr: 0.045568, batch_cost: 3.9749, reader_cost: 2.98335, ips: 64.4047 samples/sec | ETA 00:29:48
2021-11-03 00:24:41 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
/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))
380/380 [==============================] - 16s 41ms/step - batch_cost: 0.0412 - reader cost: 0.001
2021-11-03 00:24:57 [INFO]	[EVAL] #Images: 380 mIoU: 0.3135 Acc: 0.5048 Kappa: 0.1320 
2021-11-03 00:24:57 [INFO]	[EVAL] Class IoU: 
[0.4423 0.1847]
2021-11-03 00:24:57 [INFO]	[EVAL] Class Acc: 
[0.449  0.8941]
2021-11-03 00:24:57 [INFO]	[EVAL] The model with the best validation mIoU (0.3135) was saved at iter 50.
2021-11-03 00:25:36 [INFO]	[TRAIN] epoch: 12, iter: 60/500, loss: 0.5969, lr: 0.044657, batch_cost: 3.9748, reader_cost: 2.99298, ips: 64.4065 samples/sec | ETA 00:29:08
2021-11-03 00:26:16 [INFO]	[TRAIN] epoch: 14, iter: 70/500, loss: 0.5787, lr: 0.043745, batch_cost: 3.9955, reader_cost: 3.00726, ips: 64.0719 samples/sec | ETA 00:28:38
2021-11-03 00:26:56 [INFO]	[TRAIN] epoch: 16, iter: 80/500, loss: 0.5651, lr: 0.042830, batch_cost: 3.9896, reader_cost: 3.00293, ips: 64.1663 samples/sec | ETA 00:27:55
2021-11-03 00:27:36 [INFO]	[TRAIN] epoch: 18, iter: 90/500, loss: 0.5672, lr: 0.041914, batch_cost: 4.0269, reader_cost: 3.04538, ips: 63.5724 samples/sec | ETA 00:27:31
2021-11-03 00:28:17 [INFO]	[TRAIN] epoch: 20, iter: 100/500, loss: 0.5724, lr: 0.040995, batch_cost: 4.0466, reader_cost: 3.06060, ips: 63.2635 samples/sec | ETA 00:26:58
2021-11-03 00:28:17 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 41ms/step - batch_cost: 0.0412 - reader cost: 0.001
2021-11-03 00:28:33 [INFO]	[EVAL] #Images: 380 mIoU: 0.4029 Acc: 0.5850 Kappa: 0.2497 
2021-11-03 00:28:33 [INFO]	[EVAL] Class IoU: 
[0.4822 0.3236]
2021-11-03 00:28:33 [INFO]	[EVAL] Class Acc: 
[0.4942 0.9107]
2021-11-03 00:28:33 [INFO]	[EVAL] The model with the best validation mIoU (0.4029) was saved at iter 100.
2021-11-03 00:29:12 [INFO]	[TRAIN] epoch: 22, iter: 110/500, loss: 0.5607, lr: 0.040073, batch_cost: 3.9727, reader_cost: 2.98274, ips: 64.4393 samples/sec | ETA 00:25:49
2021-11-03 00:29:52 [INFO]	[TRAIN] epoch: 24, iter: 120/500, loss: 0.5449, lr: 0.039150, batch_cost: 3.9292, reader_cost: 2.94368, ips: 65.1530 samples/sec | ETA 00:24:53
2021-11-03 00:30:31 [INFO]	[TRAIN] epoch: 26, iter: 130/500, loss: 0.5525, lr: 0.038224, batch_cost: 3.9716, reader_cost: 2.98407, ips: 64.4574 samples/sec | ETA 00:24:29
2021-11-03 00:31:11 [INFO]	[TRAIN] epoch: 28, iter: 140/500, loss: 0.5465, lr: 0.037295, batch_cost: 3.9985, reader_cost: 3.00955, ips: 64.0240 samples/sec | ETA 00:23:59
2021-11-03 00:31:51 [INFO]	[TRAIN] epoch: 30, iter: 150/500, loss: 0.5349, lr: 0.036364, batch_cost: 3.9796, reader_cost: 2.99245, ips: 64.3279 samples/sec | ETA 00:23:12
2021-11-03 00:31:51 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 43ms/step - batch_cost: 0.0424 - reader cost: 0.001
2021-11-03 00:32:07 [INFO]	[EVAL] #Images: 380 mIoU: 0.4709 Acc: 0.6444 Kappa: 0.3430 
2021-11-03 00:32:07 [INFO]	[EVAL] Class IoU: 
[0.5197 0.422 ]
2021-11-03 00:32:07 [INFO]	[EVAL] Class Acc: 
[0.535  0.9247]
2021-11-03 00:32:07 [INFO]	[EVAL] The model with the best validation mIoU (0.4709) was saved at iter 150.
2021-11-03 00:32:47 [INFO]	[TRAIN] epoch: 32, iter: 160/500, loss: 0.5217, lr: 0.035430, batch_cost: 3.9988, reader_cost: 3.01584, ips: 64.0188 samples/sec | ETA 00:22:39
2021-11-03 00:33:28 [INFO]	[TRAIN] epoch: 34, iter: 170/500, loss: 0.5302, lr: 0.034494, batch_cost: 4.0304, reader_cost: 3.03914, ips: 63.5179 samples/sec | ETA 00:22:10
2021-11-03 00:34:07 [INFO]	[TRAIN] epoch: 36, iter: 180/500, loss: 0.5218, lr: 0.033555, batch_cost: 3.9072, reader_cost: 2.91663, ips: 65.5197 samples/sec | ETA 00:20:50
2021-11-03 00:34:47 [INFO]	[TRAIN] epoch: 38, iter: 190/500, loss: 0.5239, lr: 0.032612, batch_cost: 3.9990, reader_cost: 3.00959, ips: 64.0163 samples/sec | ETA 00:20:39
2021-11-03 00:35:27 [INFO]	[TRAIN] epoch: 40, iter: 200/500, loss: 0.5226, lr: 0.031667, batch_cost: 3.9884, reader_cost: 3.00057, ips: 64.1854 samples/sec | ETA 00:19:56
2021-11-03 00:35:27 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 41ms/step - batch_cost: 0.0410 - reader cost: 0.00
2021-11-03 00:35:42 [INFO]	[EVAL] #Images: 380 mIoU: 0.4626 Acc: 0.6373 Kappa: 0.3317 
2021-11-03 00:35:42 [INFO]	[EVAL] Class IoU: 
[0.515  0.4102]
2021-11-03 00:35:42 [INFO]	[EVAL] Class Acc: 
[0.5298 0.9236]
2021-11-03 00:35:42 [INFO]	[EVAL] The model with the best validation mIoU (0.4709) was saved at iter 150.
2021-11-03 00:36:22 [INFO]	[TRAIN] epoch: 42, iter: 210/500, loss: 0.4973, lr: 0.030719, batch_cost: 3.9429, reader_cost: 2.95770, ips: 64.9269 samples/sec | ETA 00:19:03
2021-11-03 00:37:01 [INFO]	[TRAIN] epoch: 44, iter: 220/500, loss: 0.5054, lr: 0.029767, batch_cost: 3.9474, reader_cost: 2.95505, ips: 64.8523 samples/sec | ETA 00:18:25
2021-11-03 00:37:41 [INFO]	[TRAIN] epoch: 46, iter: 230/500, loss: 0.5073, lr: 0.028812, batch_cost: 3.9892, reader_cost: 2.99883, ips: 64.1738 samples/sec | ETA 00:17:57
2021-11-03 00:38:21 [INFO]	[TRAIN] epoch: 48, iter: 240/500, loss: 0.5048, lr: 0.027853, batch_cost: 3.9480, reader_cost: 2.96041, ips: 64.8430 samples/sec | ETA 00:17:06
2021-11-03 00:39:01 [INFO]	[TRAIN] epoch: 50, iter: 250/500, loss: 0.4908, lr: 0.026891, batch_cost: 3.9883, reader_cost: 3.00040, ips: 64.1876 samples/sec | ETA 00:16:37
2021-11-03 00:39:01 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 41ms/step - batch_cost: 0.0410 - reader cost: 0.001
2021-11-03 00:39:16 [INFO]	[EVAL] #Images: 380 mIoU: 0.4518 Acc: 0.6282 Kappa: 0.3179 
2021-11-03 00:39:16 [INFO]	[EVAL] Class IoU: 
[0.51   0.3936]
2021-11-03 00:39:16 [INFO]	[EVAL] Class Acc: 
[0.5231 0.9268]
2021-11-03 00:39:16 [INFO]	[EVAL] The model with the best validation mIoU (0.4709) was saved at iter 150.
2021-11-03 00:39:55 [INFO]	[TRAIN] epoch: 52, iter: 260/500, loss: 0.5051, lr: 0.025925, batch_cost: 3.9250, reader_cost: 2.92987, ips: 65.2232 samples/sec | ETA 00:15:41
2021-11-03 00:40:36 [INFO]	[TRAIN] epoch: 54, iter: 270/500, loss: 0.4934, lr: 0.024954, batch_cost: 4.0248, reader_cost: 3.03510, ips: 63.6059 samples/sec | ETA 00:15:25
2021-11-03 00:41:16 [INFO]	[TRAIN] epoch: 56, iter: 280/500, loss: 0.4881, lr: 0.023980, batch_cost: 3.9854, reader_cost: 2.99336, ips: 64.2341 samples/sec | ETA 00:14:36
2021-11-03 00:41:55 [INFO]	[TRAIN] epoch: 58, iter: 290/500, loss: 0.4767, lr: 0.023001, batch_cost: 3.9613, reader_cost: 2.97126, ips: 64.6258 samples/sec | ETA 00:13:51
2021-11-03 00:42:36 [INFO]	[TRAIN] epoch: 60, iter: 300/500, loss: 0.4770, lr: 0.022018, batch_cost: 4.0457, reader_cost: 3.05480, ips: 63.2773 samples/sec | ETA 00:13:29
2021-11-03 00:42:36 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 15s 41ms/step - batch_cost: 0.0405 - reader cost: 0.00
2021-11-03 00:42:51 [INFO]	[EVAL] #Images: 380 mIoU: 0.5178 Acc: 0.6839 Kappa: 0.4066 
2021-11-03 00:42:51 [INFO]	[EVAL] Class IoU: 
[0.5472 0.4884]
2021-11-03 00:42:51 [INFO]	[EVAL] Class Acc: 
[0.5666 0.9267]
2021-11-03 00:42:51 [INFO]	[EVAL] The model with the best validation mIoU (0.5178) was saved at iter 300.
2021-11-03 00:43:31 [INFO]	[TRAIN] epoch: 62, iter: 310/500, loss: 0.4763, lr: 0.021029, batch_cost: 3.9940, reader_cost: 3.00568, ips: 64.0955 samples/sec | ETA 00:12:38
2021-11-03 00:44:11 [INFO]	[TRAIN] epoch: 64, iter: 320/500, loss: 0.4769, lr: 0.020036, batch_cost: 4.0372, reader_cost: 3.04802, ips: 63.4096 samples/sec | ETA 00:12:06
2021-11-03 00:44:51 [INFO]	[TRAIN] epoch: 66, iter: 330/500, loss: 0.4745, lr: 0.019037, batch_cost: 3.9867, reader_cost: 3.00122, ips: 64.2138 samples/sec | ETA 00:11:17
2021-11-03 00:45:32 [INFO]	[TRAIN] epoch: 68, iter: 340/500, loss: 0.4768, lr: 0.018032, batch_cost: 4.0501, reader_cost: 3.06707, ips: 63.2086 samples/sec | ETA 00:10:48
2021-11-03 00:46:12 [INFO]	[TRAIN] epoch: 70, iter: 350/500, loss: 0.4729, lr: 0.017021, batch_cost: 3.9921, reader_cost: 3.00525, ips: 64.1261 samples/sec | ETA 00:09:58
2021-11-03 00:46:12 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 42ms/step - batch_cost: 0.0418 - reader cost: 0.001
2021-11-03 00:46:28 [INFO]	[EVAL] #Images: 380 mIoU: 0.5345 Acc: 0.6975 Kappa: 0.4282 
2021-11-03 00:46:28 [INFO]	[EVAL] Class IoU: 
[0.5561 0.513 ]
2021-11-03 00:46:28 [INFO]	[EVAL] Class Acc: 
[0.5791 0.9217]
2021-11-03 00:46:28 [INFO]	[EVAL] The model with the best validation mIoU (0.5345) was saved at iter 350.
2021-11-03 00:47:07 [INFO]	[TRAIN] epoch: 72, iter: 360/500, loss: 0.4644, lr: 0.016003, batch_cost: 3.9630, reader_cost: 2.97853, ips: 64.5982 samples/sec | ETA 00:09:14
2021-11-03 00:47:48 [INFO]	[TRAIN] epoch: 74, iter: 370/500, loss: 0.4519, lr: 0.014978, batch_cost: 4.0846, reader_cost: 3.09954, ips: 62.6738 samples/sec | ETA 00:08:51
2021-11-03 00:48:28 [INFO]	[TRAIN] epoch: 76, iter: 380/500, loss: 0.4556, lr: 0.013945, batch_cost: 3.9984, reader_cost: 3.00991, ips: 64.0259 samples/sec | ETA 00:07:59
2021-11-03 00:49:08 [INFO]	[TRAIN] epoch: 78, iter: 390/500, loss: 0.4544, lr: 0.012903, batch_cost: 3.9866, reader_cost: 2.99738, ips: 64.2149 samples/sec | ETA 00:07:18
2021-11-03 00:49:48 [INFO]	[TRAIN] epoch: 80, iter: 400/500, loss: 0.4501, lr: 0.011852, batch_cost: 4.0086, reader_cost: 3.02056, ips: 63.8634 samples/sec | ETA 00:06:40
2021-11-03 00:49:48 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 41ms/step - batch_cost: 0.0413 - reader cost: 0.001
2021-11-03 00:50:04 [INFO]	[EVAL] #Images: 380 mIoU: 0.4670 Acc: 0.6418 Kappa: 0.3409 
2021-11-03 00:50:04 [INFO]	[EVAL] Class IoU: 
[0.521  0.4131]
2021-11-03 00:50:04 [INFO]	[EVAL] Class Acc: 
[0.5326 0.9392]
2021-11-03 00:50:04 [INFO]	[EVAL] The model with the best validation mIoU (0.5345) was saved at iter 350.
2021-11-03 00:50:44 [INFO]	[TRAIN] epoch: 82, iter: 410/500, loss: 0.4472, lr: 0.010790, batch_cost: 4.0134, reader_cost: 3.02345, ips: 63.7862 samples/sec | ETA 00:06:01
2021-11-03 00:51:24 [INFO]	[TRAIN] epoch: 84, iter: 420/500, loss: 0.4489, lr: 0.009717, batch_cost: 3.9976, reader_cost: 3.01048, ips: 64.0384 samples/sec | ETA 00:05:19
2021-11-03 00:52:04 [INFO]	[TRAIN] epoch: 86, iter: 430/500, loss: 0.4547, lr: 0.008630, batch_cost: 3.9965, reader_cost: 3.01080, ips: 64.0565 samples/sec | ETA 00:04:39
2021-11-03 00:52:44 [INFO]	[TRAIN] epoch: 88, iter: 440/500, loss: 0.4307, lr: 0.007528, batch_cost: 3.9896, reader_cost: 3.00905, ips: 64.1663 samples/sec | ETA 00:03:59
2021-11-03 00:53:24 [INFO]	[TRAIN] epoch: 90, iter: 450/500, loss: 0.4432, lr: 0.006408, batch_cost: 3.9713, reader_cost: 2.98270, ips: 64.4622 samples/sec | ETA 00:03:18
2021-11-03 00:53:24 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 15s 41ms/step - batch_cost: 0.0406 - reader cost: 0.001
2021-11-03 00:53:39 [INFO]	[EVAL] #Images: 380 mIoU: 0.5603 Acc: 0.7185 Kappa: 0.4641 
2021-11-03 00:53:39 [INFO]	[EVAL] Class IoU: 
[0.5741 0.5465]
2021-11-03 00:53:39 [INFO]	[EVAL] Class Acc: 
[0.5981 0.9273]
2021-11-03 00:53:39 [INFO]	[EVAL] The model with the best validation mIoU (0.5603) was saved at iter 450.
2021-11-03 00:54:19 [INFO]	[TRAIN] epoch: 92, iter: 460/500, loss: 0.4341, lr: 0.005265, batch_cost: 3.9555, reader_cost: 2.97218, ips: 64.7195 samples/sec | ETA 00:02:38
2021-11-03 00:54:59 [INFO]	[TRAIN] epoch: 94, iter: 470/500, loss: 0.4358, lr: 0.004094, batch_cost: 3.9997, reader_cost: 3.01172, ips: 64.0056 samples/sec | ETA 00:01:59
2021-11-03 00:55:39 [INFO]	[TRAIN] epoch: 96, iter: 480/500, loss: 0.4367, lr: 0.002883, batch_cost: 4.0574, reader_cost: 3.07041, ips: 63.0940 samples/sec | ETA 00:01:21
2021-11-03 00:56:20 [INFO]	[TRAIN] epoch: 98, iter: 490/500, loss: 0.4348, lr: 0.001611, batch_cost: 4.0403, reader_cost: 3.05264, ips: 63.3610 samples/sec | ETA 00:00:40
2021-11-03 00:57:00 [INFO]	[TRAIN] epoch: 100, iter: 500/500, loss: 0.4246, lr: 0.000186, batch_cost: 4.0148, reader_cost: 3.02799, ips: 63.7634 samples/sec | ETA 00:00:00
2021-11-03 00:57:00 [INFO]	Start evaluating (total_samples: 380, total_iters: 380)...
380/380 [==============================] - 16s 41ms/step - batch_cost: 0.0409 - reader cost: 0.00
2021-11-03 00:57:15 [INFO]	[EVAL] #Images: 380 mIoU: 0.5772 Acc: 0.7321 Kappa: 0.4874 
2021-11-03 00:57:15 [INFO]	[EVAL] Class IoU: 
[0.586  0.5684]
2021-11-03 00:57:15 [INFO]	[EVAL] Class Acc: 
[0.6112 0.9296]
2021-11-03 00:57:16 [INFO]	[EVAL] The model with the best validation mIoU (0.5772) was saved at iter 500.
<class 'paddle.nn.layer.conv.Conv2D'>'s flops has been counted
<class 'paddle.nn.layer.norm.BatchNorm2D'>'s flops has been counted
/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: 795137504     Total Params: 1543954
Copy the code

Five, the test

! python val.py \ --config /home/aistudio/bisenet_optic_disc_512x512_1k.yml\ --model_path output/best_model/model.pdparamsCopy the code

2021-11-03 00:59:45 [INFO] ---------------Config Information--------------- batch_size: 256 iters: 500 loss: coef: -1 types: - type: CrossEntropyLoss LR_scheduler: end_LR: 0 Learning_rate: 0.05 Power: 0.9 Type: theoreialdecay model: backbone: align_corners: false type: HRNet_W18_Small_V1 num_classes: 2 pretrained: null type: FCN optimizer: momentum: 0.9 type: SGD weight_decay: 0.0005 train_dataset: dataset_root: /home/aistudio/cat/ mode: train num_classes: 2 train_path: / home/aistudio/cat/train. TXT transforms: - max_scale_factor: min_scale_factor 2.0:0.5 scale_step_size: 0.25 type: ResizeStepScaling - crop_size: -224-224 type: randompaddingcorp-type: RandomHorizontalFlip - brightness_range: 0.4 Contrast_range: 0.4 saturation_range: 0.4 type: RandomDistort - type: Normalize type: Dataset val_dataset: dataset_root: /home/aistudio/cat/ mode: val num_classes: 2 transforms: - type: Normalize type: Dataset val_path: / home/aistudio/cat/eval. TXT -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- W1103 00:59:45. 931428, 9288 Device_context. cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: Cc :422] Device: 0, cuDNN Version: 2021-11-03 00:59:55 [INFO] Loading Pretrained Model from output/ Best_model/model.pdParams 2021-11-03 00:59:57 [INFO] There are 363/363 variables loaded into FCN. 2021-11-03 00:59:57 [INFO] Loaded trained params of model successfully 2021-11-03 00:59:57 [INFO] Start evaluating (total_samples: 380, total_iters: 380)... / 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, 380/380 [lhs_dtype)) = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =] - 16 s 41 ms/step - batch_cost: 0.0409 - reader cost: 0.002 2021-11-03 01:00:13 [INFO] [EVAL] #Images: 380 mIoU: 0.5772 Acc: 0.7321 Kappa: 2021-11-03 01:00:13 [INFO] [EVAL] Class IoU: [0.586 0.5684] 2021-11-03 01:00:13 [INFO] [EVAL] Class Acc: [0.6112 0.9296]Copy the code

6. Export the static model

! python export.py \ --config /home/aistudio/bisenet_optic_disc_512x512_1k.yml\ --model_path output/best_model/model.pdparamsCopy the code

W1103 01:00:57.781003  9593 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 10.1, Runtime API Version: 10.1
W1103 01:00:57.785893  9593 device_context.cc:422] device: 0, cuDNN Version: 7.6.
2021-11-03 01:01:09 [INFO]	Loaded trained params of model successfully.
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/utils.py:77: DeprecationWarning: Using or importing the ABCs from 'collections' instead of from 'collections.abc' is deprecated, and in 3.8 it will stop working
  return (isinstance(seq, collections.Sequence) and
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpiu3mi0oc.py:28
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpn_6sfwjx.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp2eo8_6gd.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpwc3aqih6.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp3i52fv6g.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp331woke4.py:36
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp331woke4.py:58
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpt_4sfeo_.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp64kisiuw.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpv305ofe5.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpo4s_zth8.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpba82b2d2.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmps_4r1xg0.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp_572s8qy.py:36
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp_572s8qy.py:58
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpyaadsihk.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp0vdk9o9u.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpiydq81o7.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmppw5nb4yl.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmpm9qyd52p.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmprgan86ed.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp5f_8uade.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp7ye2d18e.py:27
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp_l3u6xjv.py:36
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/math_op_patch.py:322: UserWarning: /tmp/tmp_l3u6xjv.py:58
The behavior of expression A + B has been unified with elementwise_add(X, Y, axis=-1) from Paddle 2.0. If your code works well in the older versions but crashes in this version, try to use elementwise_add(X, Y, axis=0) instead of A + B. This transitional warning will be dropped in the future.
  op_type, op_type, EXPRESSION_MAP[method_name]))
2021-11-03 01:01:11 [INFO]	Model is saved in ./output.
Copy the code

Seven, forecasting

deploy.yaml

Deploy:
  model: model.pdmodel
  params: model.pdiparams
  transforms:
  - type: Normalize
Copy the code

! pip install -e .Copy the code

from PIL import Image
img=Image.open('/home/aistudio/cat/images/Siamese_79.jpg')
img
Copy the code

from PIL import Image
img=Image.open('/home/aistudio/PaddleSeg/output/Siamese_79.png')
img
Copy the code

! python deploy/python/infer.py --config output/deploy.yaml --image_path /home/aistudio/cat/images/Siamese_79.jpgCopy the code

Copy the code

mo4tech.com (Moment For Technology) is a global community with thousands techies from across the global hang out!Passionate technologists, be it gadget freaks, tech enthusiasts, coders, technopreneurs, or CIOs, you would find them all here.

PaddleSeg based cat master and child image segmentation

I. Paddleseg-based cat master and child image segmentation

1. PaddleSeg profile

features

2. Data set making

3. Train custom data sets

3.1 File Structure

3.2 Contents of cat.txt:

3. Data viewing

4. Label processing

2. Data set preprocessing

Three, configuration,

Fourth, training

Five, the test

6. Export the static model

Seven, forecasting

PaddleSeg based cat master and child image segmentation

I. Paddleseg-based cat master and child image segmentation

1. PaddleSeg profile

features

2. Data set making

3. Train custom data sets

3.1 File Structure

3.2 Contents of cat.txt:

3. Data viewing

4. Label processing

2. Data set preprocessing

Three, configuration,

Fourth, training

Five, the test

6. Export the static model

Seven, forecasting

Related Posts

Introduction to Reinforcement Learning 5 – this article introduces you to DQN

Pyenv using Python matplotlib fixed

Century-old photo repair algorithm for parents with high appearance levels