From dbdcf5e5f9631ebb2076be1d0b97f65c679d837a Mon Sep 17 00:00:00 2001 From: LDOUBLEV Date: Mon, 6 Sep 2021 17:49:19 +0800 Subject: [PATCH] update detection md --- doc/doc_ch/detection.md | 52 +++++++++++++++++++++++++++++--------- doc/doc_en/detection_en.md | 42 ++++++++++++++++++++++-------- 2 files changed, 72 insertions(+), 22 deletions(-) diff --git a/doc/doc_ch/detection.md b/doc/doc_ch/detection.md index e9aaf1be..0a8188ed 100644 --- a/doc/doc_ch/detection.md +++ b/doc/doc_ch/detection.md @@ -1,10 +1,32 @@ -# 文字检测 + +# 目录 +- [1. 文字检测](#1-----) + * [1.1 数据准备](#11-----) + * [1.2 下载预训练模型](#12--------) + * [1.3 启动训练](#13-----) + * [1.4 断点训练](#14-----) + * [1.5 更换Backbone 训练](#15---backbone---) + * [1.6 指标评估](#16-----) + * [1.7 测试检测效果](#17-------) + * [1.8 转inference模型测试](#18--inference----) +- [2. FAQ](#2-faq) + + + +# 1. 文字检测 本节以icdar2015数据集为例，介绍PaddleOCR中检测模型训练、评估、测试的使用方式。 -## 数据准备 + +## 1.1 数据准备 icdar2015数据集可以从[官网](https://rrc.cvc.uab.es/?ch=4&com=downloads)下载到，首次下载需注册。 +注册完成登陆后，下载下图中红色框标出的部分，其中， `Training Set Images`下载的内容保存为`icdar_c4_train_imgs`文件夹下，`Test Set Images` 下载的内容保存为`ch4_test_images`文件夹下 + +

+ +

+ 将下载到的数据集解压到工作目录下，假设解压在 PaddleOCR/train_data/ 下。另外，PaddleOCR将零散的标注文件整理成单独的标注文件 ，您可以通过wget的方式进行下载。 ```shell @@ -42,7 +64,8 @@ json.dumps编码前的图像标注信息是包含多个字典的list，字典中 如果您想在其他数据集上训练，可以按照上述形式构建标注文件。 -## 下载预训练模型 + +## 1.2 下载预训练模型 首先下载模型backbone的pretrain model，PaddleOCR的检测模型目前支持两种backbone，分别是MobileNetV3、ResNet_vd系列， 您可以根据需求使用[PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/release/2.0/ppcls/modeling/architectures)中的模型更换backbone， @@ -59,7 +82,8 @@ wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dyg wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_ssld_pretrained.pdparams ``` -## 启动训练 + +## 1.3 启动训练 *如果您安装的是cpu版本，请将配置文件中的 `use_gpu` 字段修改为false* @@ -81,7 +105,8 @@ python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/ python3 tools/train.py -c configs/det/det_mv3_db.yml -o Optimizer.base_lr=0.0001 ``` -## 断点训练 + +## 1.4 断点训练 如果训练程序中断，如果希望加载训练中断的模型从而恢复训练，可以通过指定Global.checkpoints指定要加载的模型路径： ```shell @@ -90,7 +115,8 @@ python3 tools/train.py -c configs/det/det_mv3_db.yml -o Global.checkpoints=./you **注意**：`Global.checkpoints`的优先级高于`Global.pretrain_weights`的优先级，即同时指定两个参数时，优先加载`Global.checkpoints`指定的模型，如果`Global.checkpoints`指定的模型路径有误，会加载`Global.pretrain_weights`指定的模型。 -## 更换Backbone 训练 + +## 1.5 更换Backbone 训练 PaddleOCR将网络划分为四部分，分别在[ppocr/modeling](../../ppocr/modeling)下。 进入网络的数据将按照顺序(transforms->backbones-> necks->heads)依次通过这四个部分。 @@ -137,8 +163,8 @@ args1: args1 **注意**：如果要更换网络的其他模块，可以参考[文档](./add_new_algorithm.md)。 - -## 指标评估 + +## 1.6 指标评估 PaddleOCR计算三个OCR检测相关的指标，分别是：Precision、Recall、Hmean（F-Score）。 @@ -150,7 +176,8 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat * 注：`box_thresh`、`unclip_ratio`是DB后处理所需要的参数，在评估EAST模型时不需要设置 -## 测试检测效果 + +## 1.7 测试检测效果 测试单张图像的检测效果 ```shell @@ -167,7 +194,8 @@ python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./ python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy" ``` -## 转inference模型测试 + +## 1.8 转inference模型测试 inference 模型（`paddle.jit.save`保存的模型） 一般是模型训练，把模型结构和模型参数保存在文件中的固化模型，多用于预测部署场景。 @@ -189,8 +217,8 @@ python3 tools/infer/predict_det.py --det_algorithm="DB" --det_model_dir="./outpu python3 tools/infer/predict_det.py --det_algorithm="EAST" --det_model_dir="./output/det_db_inference/" --image_dir="./doc/imgs/" --use_gpu=True ``` - -## FAQ + +# 2. FAQ Q1: 训练模型转inference 模型之后预测效果不一致？ **A**：此类问题出现较多，问题多是trained model预测时候的预处理、后处理参数和inference model预测的时候的预处理、后处理参数不一致导致的。以det_mv3_db.yml配置文件训练的模型为例，训练模型、inference模型预测结果不一致问题解决方式如下： diff --git a/doc/doc_en/detection_en.md b/doc/doc_en/detection_en.md index 966c603b..8f12d42f 100644 --- a/doc/doc_en/detection_en.md +++ b/doc/doc_en/detection_en.md @@ -1,10 +1,32 @@ -# TEXT DETECTION +# CONTENT + +- [Paste Your Document In Here](#paste-your-document-in-here) +- [1. TEXT DETECTION](#1-text-detection) + * [1.1 DATA PREPARATION](#11-data-preparation) + * [1.2 DOWNLOAD PRETRAINED MODEL](#12-download-pretrained-model) + * [1.3 START TRAINING](#13-start-training) + * [1.4 LOAD TRAINED MODEL AND CONTINUE TRAINING](#14-load-trained-model-and-continue-training) + * [1.5 TRAINING WITH NEW BACKBONE](#15-training-with-new-backbone) + * [1.6 EVALUATION](#16-evaluation) + * [1.7 TEST](#17-test) + * [1.8 INFERENCE MODEL PREDICTION](#18-inference-model-prediction) +- [2. FAQ](#2-faq) + + +# 1. TEXT DETECTION This section uses the icdar2015 dataset as an example to introduce the training, evaluation, and testing of the detection model in PaddleOCR. -## DATA PREPARATION +## 1.1 DATA PREPARATION The icdar2015 dataset can be obtained from [official website](https://rrc.cvc.uab.es/?ch=4&com=downloads). Registration is required for downloading. + +After registering and logging in, download the part marked in the red box in the figure below. And, the content downloaded by `Training Set Images` should be saved as the folder `icdar_c4_train_imgs`, and the content downloaded by `Test Set Images` is saved as the folder `ch4_test_images` + +

+ +

+ Decompress the downloaded dataset to the working directory, assuming it is decompressed under PaddleOCR/train_data/. In addition, PaddleOCR organizes many scattered annotation files into two separate annotation files for train and test respectively, which can be downloaded by wget: ```shell # Under the PaddleOCR path @@ -36,7 +58,7 @@ The `points` in the dictionary represent the coordinates (x, y) of the four poin If you want to train PaddleOCR on other datasets, please build the annotation file according to the above format. -## DOWNLOAD PRETRAINED MODEL +## 1.2 DOWNLOAD PRETRAINED MODEL First download the pretrained model. The detection model of PaddleOCR currently supports 3 backbones, namely MobileNetV3, ResNet18_vd and ResNet50_vd. You can use the model in [PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/release/2.0/ppcls/modeling/architectures) to replace backbone according to your needs. And the responding download link of backbone pretrain weights can be found in (https://github.com/PaddlePaddle/PaddleClas/blob/release%2F2.0/README_cn.md#resnet%E5%8F%8A%E5%85%B6vd%E7%B3%BB%E5%88%97). @@ -52,7 +74,7 @@ wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dyg ``` -## START TRAINING +## 1.3 START TRAINING *If CPU version installed, please set the parameter `use_gpu` to `false` in the configuration.* ```shell python3 tools/train.py -c configs/det/det_mv3_db.yml \ @@ -75,7 +97,7 @@ python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs ``` -## LOAD TRAINED MODEL AND CONTINUE TRAINING +## 1.4 LOAD TRAINED MODEL AND CONTINUE TRAINING If you expect to load trained model and continue the training again, you can specify the parameter `Global.checkpoints` as the model path to be loaded. For example: @@ -86,7 +108,7 @@ python3 tools/train.py -c configs/det/det_mv3_db.yml -o Global.checkpoints=./you **Note**: The priority of `Global.checkpoints` is higher than that of `Global.pretrain_weights`, that is, when two parameters are specified at the same time, the model specified by `Global.checkpoints` will be loaded first. If the model path specified by `Global.checkpoints` is wrong, the one specified by `Global.pretrain_weights` will be loaded. -## TRAINING WITH NEW BACKBONE +## 1.5 TRAINING WITH NEW BACKBONE The network part completes the construction of the network, and PaddleOCR divides the network into four parts, which are under [ppocr/modeling](../../ppocr/modeling). The data entering the network will pass through these four parts in sequence(transforms->backbones-> necks->heads). @@ -136,7 +158,7 @@ After adding the four-part modules of the network, you only need to configure th **NOTE**: More details about replace Backbone and other mudule can be found in [doc](add_new_algorithm_en.md). -## EVALUATION +## 1.6 EVALUATION PaddleOCR calculates three indicators for evaluating performance of OCR detection task: Precision, Recall, and Hmean(F-Score). @@ -151,7 +173,7 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat * Note: `box_thresh` and `unclip_ratio` are parameters required for DB post-processing, and not need to be set when evaluating the EAST and SAST model. -## TEST +## 1.7 TEST Test the detection result on a single image: ```shell @@ -169,7 +191,7 @@ Test the detection result on all images in the folder: python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy" ``` -## INFERENCE MODEL PREDICTION +## 1.8 INFERENCE MODEL PREDICTION The inference model (the model saved by `paddle.jit.save`) is generally a solidified model saved after the model training is completed, and is mostly used to give prediction in deployment. @@ -192,7 +214,7 @@ If it is other detection algorithms, such as the EAST, the det_algorithm paramet python3 tools/infer/predict_det.py --det_algorithm="EAST" --det_model_dir="./output/det_db_inference/" --image_dir="./doc/imgs/" --use_gpu=True ``` -## FAQ +# 2. FAQ Q1: The prediction results of trained model and inference model are inconsistent? **A**: Most of the problems are caused by the inconsistency of the pre-processing and post-processing parameters during the prediction of the trained model and the pre-processing and post-processing parameters during the prediction of the inference model. Taking the model trained by the det_mv3_db.yml configuration file as an example, the solution to the problem of inconsistent prediction results between the training model and the inference model is as follows: -- GitLab