[paddlepaddle] paddedetection face recognition custom data set

【PaddlePaddle】 PaddleDetection Face recognition Custom datasets

Use paddleDetection Face recognition is realized

---

---

# Preface

Use paddleDetection Face recognition

---

brief introduction

PaddleDetection Development kit of flying propeller target detection , It aims to help developers complete the construction of detection model faster and better 、 Training 、 Optimize and deploy the whole development process .

PaddleDetection A variety of mainstream target detection algorithms are modularized , Provides a wealth of data enhancement strategies 、 Network module components （ Such as backbone network ）、 Loss function, etc , It integrates model compression and cross platform high-performance deployment capabilities .

After a long period of industrial practice ,PaddleDetection Have a smooth 、 Excellent use experience , Industrial quality inspection 、 Remote sensing image detection 、 There's no patrol 、 The new retail 、 Internet 、 It is widely used by developers in more than ten industries such as scientific research .

characteristic

• The model is rich : contain object detection 、 Instance segmentation 、 Face detection etc. 100+ Pre training models , It covers a variety of Global champion programme
• Simple to use ： Modular design , Decouple the various network components , Developers can easily build 、 Try various detection models and optimization strategies , Get high performance quickly 、 Customized Algorithm .
• Open end to end : Enhanced from data 、 networking 、 Training 、 Compress 、 Deploy end-to-end connection , And fully support Cloud / Border Multi architecture 、 Multi device deployment .
• High performance : High performance kernel based on propeller , Model training speed and video memory occupation advantages are obvious . Support FP16 Training , Support multi machine training .

One 、 Data set production

1. collecting data

Collect and run the following code from the local camera , It can also be obtained from other places

# -*- coding: utf-8 -*-
#######  Run locally !!!!!!!!!
import cv2
import os


path = "./pictures/" #  Image saving path 
if not os.path.exists(path):
    os.makedirs(path)
cap = cv2.VideoCapture(0)
i = 0
while (1):
    ret, frame = cap.read()
    k = cv2.waitKey(1)
    if k == 27:
        break
    elif k == ord('s'):
        cv2.imwrite(path + str(i) + '.jpg', frame)
        print("save" + str(i) + ".jpg")
        i += 1
    cv2.imshow("capture", frame)
cap.release()
cv2.destroyAllWindows()

2. mark

open labelimg

1. Open the picture folder

2. Click on change_save_dir choice xml Save the path

3. mark

Two 、 download

# Download propeller 
https://www.paddlepaddle.org.cn/
# download paddleDetection

# pip install paddledet
pip install paddledet==2.1.0 -i https://mirror.baidu.com/pypi/simple

#  Download and use the configuration files and code samples in the source code 
git clone https://github.com/PaddlePaddle/PaddleDetection.git
cd PaddleDetection
#  Install other dependencies 
pip install -r requirements.txt

3、 ... and 、 Training

1. Modify the configuration file

# Use ssd_mobilenet

#vim configs/ssd/ssd_mobilenet_v1_voc.yml
# Mainly modify the data set location 
TrainReader:
  inputs_def:
    image_shape: [3, 300, 300]
    fields: ['image', 'gt_bbox', 'gt_class']
  dataset:
    !VOCDataSet
    anno_path: trainval.txt
    dataset_dir: /home/aiuser/mtl/data/face_demo/VOCdevkit
    
EvalReader:
  inputs_def:
    image_shape: [3, 300, 300]
    fields: ['image', 'gt_bbox', 'gt_class', 'im_shape', 'im_id', 'is_difficult']
  dataset:
    !VOCDataSet
    dataset_dir: /home/aiuser/mtl/data/face_demo/VOCdevkit
    anno_path: val.txt
    
TestReader:
  inputs_def:
    image_shape: [3,300,300]
    fields: ['image', 'im_id', 'im_shape']
  dataset:
    !ImageFolder
    dataset_dir: /home/aiuser/mtl/data/face_demo/VOCdevkit
    anno_path: label_list.txt

Full profile

architecture: SSD
pretrain_weights: https://paddlemodels.bj.bcebos.com/object_detection/ssd_mobilenet_v1_coco_pretrained.tar
use_gpu: true
max_iters: 2800
snapshot_iter: 2000
log_iter: 1
metric: VOC
map_type: 11point
save_dir: output
weights: output/ssd_mobilenet_v1_voc/model_final
# 20(label_class) + 1(background)
num_classes: 25

SSD:
  backbone: MobileNet
  multi_box_head: MultiBoxHead
  output_decoder:
    background_label: 0
    keep_top_k: 200
    nms_eta: 1.0
    nms_threshold: 0.45
    nms_top_k: 400
    score_threshold: 0.01

MobileNet:
  norm_decay: 0.
  conv_group_scale: 1
  conv_learning_rate: 0.1
  extra_block_filters: [[256, 512], [128, 256], [128, 256], [64, 128]]
  with_extra_blocks: true

MultiBoxHead:
  aspect_ratios: [[2.], [2., 3.], [2., 3.], [2., 3.], [2., 3.], [2., 3.]]
  base_size: 300
  flip: true
  max_ratio: 90
  max_sizes: [[], 150.0, 195.0, 240.0, 285.0, 300.0]
  min_ratio: 20
  min_sizes: [60.0, 105.0, 150.0, 195.0, 240.0, 285.0]
  offset: 0.5

LearningRate:
  schedulers:
  - !PiecewiseDecay
    milestones: [10000, 15000, 20000, 25000]
    values: [0.001, 0.0005, 0.00025, 0.0001, 0.00001]

OptimizerBuilder:
  optimizer:
    momentum: 0.0
    type: RMSPropOptimizer
  regularizer:
    factor: 0.00005
    type: L2

TrainReader:
  inputs_def:
    image_shape: [3, 300, 300]
    fields: ['image', 'gt_bbox', 'gt_class']
  dataset:
    !VOCDataSet
    anno_path: trainval.txt
    dataset_dir: /home/aiuser/mtl/data/face_demo/VOCdevkit
    use_default_label: false
  sample_transforms:
  - !DecodeImage
    to_rgb: true
  - !RandomDistort
    brightness_lower: 0.875
    brightness_upper: 1.125
    is_order: true
  - !RandomExpand
    fill_value: [127.5, 127.5, 127.5]
  - !RandomCrop
    allow_no_crop: false
  - !NormalizeBox {}
  - !ResizeImage
    interp: 1
    target_size: 300
    use_cv2: false
  - !RandomFlipImage
    is_normalized: true
  - !Permute {}
  - !NormalizeImage
    is_scale: false
    mean: [127.5, 127.5, 127.5]
    std: [127.502231, 127.502231, 127.502231]
  batch_size: 4
  shuffle: true
  drop_last: true
  worker_num: 8
  bufsize: 16
  use_process: true

EvalReader:
  inputs_def:
    image_shape: [3, 300, 300]
    fields: ['image', 'gt_bbox', 'gt_class', 'im_shape', 'im_id', 'is_difficult']
  dataset:
    !VOCDataSet
    dataset_dir: /home/aiuser/mtl/data/face_demo/VOCdevkit
    anno_path: val.txt
    use_default_label: false
  sample_transforms:
  - !DecodeImage
    to_rgb: true
  - !NormalizeBox {}
  - !ResizeImage
    interp: 1
    target_size: 300
    use_cv2: false
  - !Permute {}
  - !NormalizeImage
    is_scale: false
    mean: [127.5, 127.5, 127.5]
    std: [127.502231, 127.502231, 127.502231]
  batch_size: 4
  worker_num: 8
  bufsize: 16
  use_process: false


TestReader:
  inputs_def:
    image_shape: [3,300,300]
    fields: ['image', 'im_id', 'im_shape']
  dataset:
    !ImageFolder
    dataset_dir: /home/aiuser/mtl/data/face_demo/VOCdevkit
    anno_path: label_list.txt
    use_default_label: false
  sample_transforms:
  - !DecodeImage
    to_rgb: true
  - !ResizeImage
    interp: 1
    max_size: 0
    target_size: 300
    use_cv2: true
  - !Permute {}
  - !NormalizeImage
    is_scale: false
    mean: [127.5, 127.5, 127.5]
    std: [127.502231, 127.502231, 127.502231]
  batch_size: 1

2. Start training

python -u tools/train.py -c configs/ssd/ssd_mobilenet_v1_voc.yml -o --eval

[ Failed to transfer the external chain picture , The origin station may have anti-theft chain mechanism , It is suggested to save the pictures and upload them directly (img-qhhnrgL3-1656931184790)(C:\Users\e9\AppData\Roaming\Typora\typora-user-images\image-20210712141658919.png)]

3. The export model

python  tools/export_model.py -c configs/ssd/ssd_mobilenet_v1_voc.yml --output_dir=./inference_model

Four 、 Use

python -u deploy/python/infer.py --model_dir D:/Paddle/PaddleDetection/PaddleDetection/output/ssd_mobilenet_v1_voc/ssd_mobilenet_v1_voc

---

nfigs/ssd/ssd_mobilenet_v1_voc.yml --output_dir=./inference_model ```

Four 、 Use

python -u deploy/python/infer.py --model_dir D:/Paddle/PaddleDetection/PaddleDetection/output/ssd_mobilenet_v1_voc/ssd_mobilenet_v1_voc

---

# summary