Merge remote-tracking branch 'origin/dygraph' into dygraph

benchmark/readme.md

 
-# PaddleOCR DB/EAST 算法训练benchmark测试
+# PaddleOCR DB/EAST/PSE 算法训练benchmark测试
 
 PaddleOCR/benchmark目录下的文件用于获取并分析训练日志。
 训练采用icdar2015数据集，包括1000张训练图像和500张测试图像。模型配置采用resnet18_vd作为backbone，分别训练batch_size=8和batch_size=16的情况。
@@ -18,7 +18,7 @@ run_det.sh 执行方式如下:
 
 ```
 # cd PaddleOCR/
-bash benchmark/run_det.sh 
+bash benchmark/run_det.sh
 ```
 
 以DB为例，将得到四个日志文件，如下：
@@ -28,7 +28,3 @@ det_res18_db_v2.0_sp_bs8_fp32_1
 det_res18_db_v2.0_mp_bs16_fp32_1
 det_res18_db_v2.0_mp_bs8_fp32_1
 ```
-
-
-
-

benchmark/run_benchmark_det.sh

@@ -6,7 +6,7 @@ function _set_params(){
     run_mode=${1:-"sp"}          # 单卡sp|多卡mp
     batch_size=${2:-"64"}
     fp_item=${3:-"fp32"}        # fp32|fp16
-    max_iter=${4:-"500"}       # 可选，如果需要修改代码提前中断
+    max_iter=${4:-"10"}       # 可选，如果需要修改代码提前中断
     model_name=${5:-"model_name"}
     run_log_path=${TRAIN_LOG_DIR:-$(pwd)}  # TRAIN_LOG_DIR 后续QA设置该参数
 
@@ -20,7 +20,7 @@ function _train(){
     echo "Train on ${num_gpu_devices} GPUs"
     echo "current CUDA_VISIBLE_DEVICES=$CUDA_VISIBLE_DEVICES, gpus=$num_gpu_devices, batch_size=$batch_size"
 
-    train_cmd="-c configs/det/${model_name}.yml -o Train.loader.batch_size_per_card=${batch_size} Global.epoch_num=${max_iter} "   
+    train_cmd="-c configs/det/${model_name}.yml -o Train.loader.batch_size_per_card=${batch_size} Global.epoch_num=${max_iter} Global.eval_batch_step=[0,20000] Global.print_batch_step=2"   
     case ${run_mode} in
       sp) 
         train_cmd="python3.7 tools/train.py "${train_cmd}""
@@ -39,18 +39,24 @@ function _train(){
         echo -e "${model_name}, SUCCESS"
         export job_fail_flag=0
     fi
-    kill -9 `ps -ef|grep 'python3.7'|awk '{print $2}'`
 
     if [ $run_mode = "mp" -a -d mylog ]; then
         rm ${log_file}
         cp mylog/workerlog.0 ${log_file}
     fi
+}
 
-    # run log analysis
-    analysis_cmd="python3.7 benchmark/analysis.py --filename ${log_file}  --mission_name ${model_name} --run_mode ${mode} --direction_id 0 --keyword 'ips:' --base_batch_size ${batch_szie} --skip_steps 1 --gpu_num ${num_gpu_devices}  --index 1  --model_mode=-1  --ips_unit=samples/sec"
+function _analysis_log(){
+    analysis_cmd="python3.7 benchmark/analysis.py --filename ${log_file}  --mission_name ${model_name} --run_mode ${run_mode} --direction_id 0 --keyword 'ips:' --base_batch_size ${batch_size} --skip_steps 1 --gpu_num ${num_gpu_devices}  --index 1  --model_mode=-1  --ips_unit=samples/sec"
     eval $analysis_cmd
 }
 
+function _kill_process(){
+    kill -9 `ps -ef|grep 'python3.7'|awk '{print $2}'`
+}
+
+
 _set_params $@
 _train
-
+_analysis_log
+_kill_process
\ No newline at end of file

benchmark/run_det.sh

@@ -3,11 +3,11 @@
 # 1 安装该模型需要的依赖 (如需开启优化策略请注明)
 python3.7 -m pip install -r requirements.txt
 # 2 拷贝该模型需要数据、预训练模型
-wget -c  -p ./tain_data/  https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015.tar && cd train_data  && tar xf icdar2015.tar && cd ../
-wget -c -p ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_pretrained.pdparams
+wget -P ./train_data/  https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015.tar && cd train_data  && tar xf icdar2015.tar && cd ../
+wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_pretrained.pdparams
 # 3 批量运行（如不方便批量，1，2需放到单个模型中）
 
-model_mode_list=(det_res18_db_v2.0 det_r50_vd_east)
+model_mode_list=(det_res18_db_v2.0 det_r50_vd_east det_r50_vd_pse)
 fp_item_list=(fp32)
 bs_list=(8 16)
 for model_mode in ${model_mode_list[@]}; do
@@ -15,11 +15,11 @@ for model_mode in ${model_mode_list[@]}; do
           for bs_item in ${bs_list[@]}; do
             echo "index is speed, 1gpus, begin, ${model_name}"
             run_mode=sp
-            CUDA_VISIBLE_DEVICES=0 bash benchmark/run_benchmark_det.sh ${run_mode} ${bs_item} ${fp_item} 10 ${model_mode}     #  (5min)
+            CUDA_VISIBLE_DEVICES=0 bash benchmark/run_benchmark_det.sh ${run_mode} ${bs_item} ${fp_item} 2 ${model_mode}     #  (5min)
             sleep 60
             echo "index is speed, 8gpus, run_mode is multi_process, begin, ${model_name}"
             run_mode=mp
-            CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 bash benchmark/run_benchmark_det.sh ${run_mode} ${bs_item} ${fp_item} 10 ${model_mode} 
+            CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 bash benchmark/run_benchmark_det.sh ${run_mode} ${bs_item} ${fp_item} 2 ${model_mode} 
             sleep 60
             done
       done

configs/rec/rec_mtb_nrtr.yml

@@ -17,7 +17,7 @@ Global:
   character_dict_path: ppocr/utils/EN_symbol_dict.txt
   max_text_length: 25
   infer_mode: False
-  use_space_char: True
+  use_space_char: False
   save_res_path: ./output/rec/predicts_nrtr.txt
 
 Optimizer:

deploy/lite/ocr_db_crnn.cc

@@ -12,12 +12,14 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-#include "paddle_api.h" // NOLINT
 #include <chrono>
+#include "paddle_api.h" // NOLINT
+#include "paddle_place.h"
 
 #include "cls_process.h"
 #include "crnn_process.h"
 #include "db_post_process.h"
+#include "AutoLog/auto_log/lite_autolog.h"
 
 using namespace paddle::lite_api; // NOLINT
 using namespace std;
@@ -27,7 +29,7 @@ void NeonMeanScale(const float *din, float *dout, int size,
                    const std::vector<float> mean,
                    const std::vector<float> scale) {
   if (mean.size() != 3 || scale.size() != 3) {
-    std::cerr << "[ERROR] mean or scale size must equal to 3\n";
+    std::cerr << "[ERROR] mean or scale size must equal to 3" << std::endl;
     exit(1);
   }
   float32x4_t vmean0 = vdupq_n_f32(mean[0]);
@@ -159,7 +161,8 @@ void RunRecModel(std::vector<std::vector<std::vector<int>>> boxes, cv::Mat img,
                  std::vector<float> &rec_text_score,
                  std::vector<std::string> charactor_dict,
                  std::shared_ptr<PaddlePredictor> predictor_cls,
-                 int use_direction_classify) {
+                 int use_direction_classify,
+                 std::vector<double> *times) {
   std::vector<float> mean = {0.5f, 0.5f, 0.5f};
   std::vector<float> scale = {1 / 0.5f, 1 / 0.5f, 1 / 0.5f};
 
@@ -226,14 +229,15 @@ void RunRecModel(std::vector<std::vector<std::vector<int>>> boxes, cv::Mat img,
 
 std::vector<std::vector<std::vector<int>>>
 RunDetModel(std::shared_ptr<PaddlePredictor> predictor, cv::Mat img,
-            std::map<std::string, double> Config) {
+            std::map<std::string, double> Config, std::vector<double> *times) {
   // Read img
   int max_side_len = int(Config["max_side_len"]);
   int det_db_use_dilate = int(Config["det_db_use_dilate"]);
 
   cv::Mat srcimg;
   img.copyTo(srcimg);
-
+  
+  auto preprocess_start = std::chrono::steady_clock::now();
   std::vector<float> ratio_hw;
   img = DetResizeImg(img, max_side_len, ratio_hw);
   cv::Mat img_fp;
@@ -248,8 +252,10 @@ RunDetModel(std::shared_ptr<PaddlePredictor> predictor, cv::Mat img,
   std::vector<float> scale = {1 / 0.229f, 1 / 0.224f, 1 / 0.225f};
   const float *dimg = reinterpret_cast<const float *>(img_fp.data);
   NeonMeanScale(dimg, data0, img_fp.rows * img_fp.cols, mean, scale);
+  auto preprocess_end = std::chrono::steady_clock::now();
 
   // Run predictor
+  auto inference_start = std::chrono::steady_clock::now();
   predictor->Run();
 
   // Get output and post process
@@ -257,8 +263,10 @@ RunDetModel(std::shared_ptr<PaddlePredictor> predictor, cv::Mat img,
       std::move(predictor->GetOutput(0)));
   auto *outptr = output_tensor->data<float>();
   auto shape_out = output_tensor->shape();
+  auto inference_end = std::chrono::steady_clock::now();
 
   // Save output
+  auto postprocess_start = std::chrono::steady_clock::now();
   float pred[shape_out[2] * shape_out[3]];
   unsigned char cbuf[shape_out[2] * shape_out[3]];
 
@@ -287,14 +295,35 @@ RunDetModel(std::shared_ptr<PaddlePredictor> predictor, cv::Mat img,
 
   std::vector<std::vector<std::vector<int>>> filter_boxes =
       FilterTagDetRes(boxes, ratio_hw[0], ratio_hw[1], srcimg);
+  auto postprocess_end = std::chrono::steady_clock::now();
+
+  std::chrono::duration<float> preprocess_diff = preprocess_end - preprocess_start;
+  times->push_back(double(preprocess_diff.count() * 1000));
+  std::chrono::duration<float> inference_diff = inference_end - inference_start;
+  times->push_back(double(inference_diff.count() * 1000));
+  std::chrono::duration<float> postprocess_diff = postprocess_end - postprocess_start;
+  times->push_back(double(postprocess_diff.count() * 1000));
 
   return filter_boxes;
 }
 
-std::shared_ptr<PaddlePredictor> loadModel(std::string model_file) {
+std::shared_ptr<PaddlePredictor> loadModel(std::string model_file, std::string power_mode, int num_threads) {
   MobileConfig config;
   config.set_model_from_file(model_file);
 
+  if (power_mode == "LITE_POWER_HIGH"){
+      config.set_power_mode(LITE_POWER_HIGH);
+  } else {
+      if (power_mode == "LITE_POWER_LOW") {
+          config.set_power_mode(LITE_POWER_HIGH);
+      } else {
+          std::cerr << "Only support LITE_POWER_HIGH or LITE_POWER_HIGH." << std::endl;
+          exit(1);
+      }
+  }
+
+  config.set_threads(num_threads);
+
   std::shared_ptr<PaddlePredictor> predictor =
       CreatePaddlePredictor<MobileConfig>(config);
   return predictor;
@@ -354,60 +383,255 @@ std::map<std::string, double> LoadConfigTxt(std::string config_path) {
   return dict;
 }
 
-int main(int argc, char **argv) {
-  if (argc < 5) {
-    std::cerr << "[ERROR] usage: " << argv[0]
-              << " det_model_file cls_model_file rec_model_file image_path "
-                 "charactor_dict\n";
+void check_params(int argc, char **argv) {
+  if (argc<=1 || (strcmp(argv[1], "det")!=0 && strcmp(argv[1], "rec")!=0 && strcmp(argv[1], "system")!=0)) {
+    std::cerr << "Please choose one mode of [det, rec, system] !" << std::endl;
     exit(1);
   }
-  std::string det_model_file = argv[1];
-  std::string rec_model_file = argv[2];
-  std::string cls_model_file = argv[3];
-  std::string img_path = argv[4];
-  std::string dict_path = argv[5];
+  if (strcmp(argv[1], "det") == 0) {
+      if (argc < 9){
+        std::cerr << "[ERROR] usage:" << argv[0]
+                  << " det det_model num_threads batchsize power_mode img_dir det_config lite_benchmark_value" << std::endl;
+        exit(1);
+      }
+  }
+
+  if (strcmp(argv[1], "rec") == 0) {
+      if (argc < 9){
+        std::cerr << "[ERROR] usage:" << argv[0]
+                  << " rec rec_model num_threads batchsize power_mode img_dir key_txt lite_benchmark_value" << std::endl;
+        exit(1);
+      }
+  }
+
+  if (strcmp(argv[1], "system") == 0) {
+      if (argc < 12){
+        std::cerr << "[ERROR] usage:" << argv[0]
+                  << " system det_model rec_model clas_model num_threads batchsize power_mode img_dir det_config key_txt lite_benchmark_value" << std::endl;
+        exit(1);
+      }
+  }
+}
+
+void system(char **argv){
+  std::string det_model_file = argv[2];
+  std::string rec_model_file = argv[3];
+  std::string cls_model_file = argv[4];
+  std::string precision = argv[5];
+  std::string num_threads = argv[6];
+  std::string batchsize = argv[7];
+  std::string power_mode = argv[8];
+  std::string img_dir = argv[9];
+  std::string det_config_path = argv[10];
+  std::string dict_path = argv[11];
+
+  if (strcmp(argv[5], "FP32") != 0 && strcmp(argv[5], "INT8") != 0) {
+      std::cerr << "Only support FP32 or INT8." << std::endl;
+      exit(1);
+  }
+
+  std::vector<cv::String> cv_all_img_names;
+  cv::glob(img_dir, cv_all_img_names);
 
   //// load config from txt file
-  auto Config = LoadConfigTxt("./config.txt");
+  auto Config = LoadConfigTxt(det_config_path);
   int use_direction_classify = int(Config["use_direction_classify"]);
 
-  auto start = std::chrono::system_clock::now();
+  auto charactor_dict = ReadDict(dict_path);
+  charactor_dict.insert(charactor_dict.begin(), "#"); // blank char for ctc
+  charactor_dict.push_back(" ");
+
+  auto det_predictor = loadModel(det_model_file, power_mode, std::stoi(num_threads));
+  auto rec_predictor = loadModel(rec_model_file, power_mode, std::stoi(num_threads));
+  auto cls_predictor = loadModel(cls_model_file, power_mode, std::stoi(num_threads));
 
-  auto det_predictor = loadModel(det_model_file);
-  auto rec_predictor = loadModel(rec_model_file);
-  auto cls_predictor = loadModel(cls_model_file);
+  for (int i = 0; i < cv_all_img_names.size(); ++i) {
+    std::cout << "The predict img: " << cv_all_img_names[i] << std::endl;
+    cv::Mat srcimg = cv::imread(cv_all_img_names[i], cv::IMREAD_COLOR);
+
+    if (!srcimg.data) {
+      std::cerr << "[ERROR] image read failed! image path: " << cv_all_img_names[i] << std::endl;
+      exit(1);
+    }
+
+    std::vector<double> det_times;
+    auto boxes = RunDetModel(det_predictor, srcimg, Config, &det_times);
+  
+    std::vector<std::string> rec_text;
+    std::vector<float> rec_text_score;
+  
+    std::vector<double> rec_times;
+    RunRecModel(boxes, srcimg, rec_predictor, rec_text, rec_text_score,
+                charactor_dict, cls_predictor, use_direction_classify, &rec_times);
+  
+    //// visualization
+    auto img_vis = Visualization(srcimg, boxes);
+  
+    //// print recognized text
+    for (int i = 0; i < rec_text.size(); i++) {
+      std::cout << i << "\t" << rec_text[i] << "\t" << rec_text_score[i]
+                << std::endl;
+    }
+  }
+}
+
+void det(int argc, char **argv) {
+  std::string det_model_file = argv[2];
+  std::string precision = argv[3];
+  std::string num_threads = argv[4];
+  std::string batchsize = argv[5];
+  std::string power_mode = argv[6];
+  std::string img_dir = argv[7];
+  std::string det_config_path = argv[8];
+
+  if (strcmp(argv[3], "FP32") != 0 && strcmp(argv[3], "INT8") != 0) {
+      std::cerr << "Only support FP32 or INT8." << std::endl;
+      exit(1);
+  }
+
+  std::vector<cv::String> cv_all_img_names;
+  cv::glob(img_dir, cv_all_img_names);
+
+  //// load config from txt file
+  auto Config = LoadConfigTxt(det_config_path);
+
+  auto det_predictor = loadModel(det_model_file, power_mode, std::stoi(num_threads));
+
+  std::vector<double> time_info = {0, 0, 0};
+  for (int i = 0; i < cv_all_img_names.size(); ++i) {
+    std::cout << "The predict img: " << cv_all_img_names[i] << std::endl;
+    cv::Mat srcimg = cv::imread(cv_all_img_names[i], cv::IMREAD_COLOR);
+
+    if (!srcimg.data) {
+      std::cerr << "[ERROR] image read failed! image path: " << cv_all_img_names[i] << std::endl;
+      exit(1);
+    }
+
+    std::vector<double> times;
+    auto boxes = RunDetModel(det_predictor, srcimg, Config, &times);
+
+    //// visualization
+    auto img_vis = Visualization(srcimg, boxes);
+    std::cout << boxes.size() << " bboxes have detected:" << std::endl;
+
+    // for (int i=0; i<boxes.size(); i++){
+    //   std::cout << "The " << i << " box:" << std::endl;
+    //   for (int j=0; j<4; j++){
+    //     for (int k=0; k<2; k++){
+    //       std::cout << boxes[i][j][k] << "\t";
+    //     }
+    //   }
+    //   std::cout << std::endl;
+    // }
+    time_info[0] += times[0];
+    time_info[1] += times[1];
+    time_info[2] += times[2];
+  }
+
+  if (strcmp(argv[9], "True") == 0) {
+    AutoLogger autolog(det_model_file, 
+                       0,
+                       0,
+                       0,
+                       std::stoi(num_threads),
+                       std::stoi(batchsize), 
+                       "dynamic", 
+                       precision, 
+                       power_mode,
+                       time_info, 
+                       cv_all_img_names.size());
+    autolog.report();
+  }
+}
+
+void rec(int argc, char **argv) {
+  std::string rec_model_file = argv[2];
+  std::string precision = argv[3];
+  std::string num_threads = argv[4];
+  std::string batchsize = argv[5];
+  std::string power_mode = argv[6];
+  std::string img_dir = argv[7];
+  std::string dict_path = argv[8];
+
+  if (strcmp(argv[3], "FP32") != 0 && strcmp(argv[3], "INT8") != 0) {
+      std::cerr << "Only support FP32 or INT8." << std::endl;
+      exit(1);
+  }
+
+  std::vector<cv::String> cv_all_img_names;
+  cv::glob(img_dir, cv_all_img_names);
 
   auto charactor_dict = ReadDict(dict_path);
   charactor_dict.insert(charactor_dict.begin(), "#"); // blank char for ctc
   charactor_dict.push_back(" ");
 
-  cv::Mat srcimg = cv::imread(img_path, cv::IMREAD_COLOR);
-  auto boxes = RunDetModel(det_predictor, srcimg, Config);
+  auto rec_predictor = loadModel(rec_model_file, power_mode, std::stoi(num_threads));
 
-  std::vector<std::string> rec_text;
-  std::vector<float> rec_text_score;
+  std::shared_ptr<PaddlePredictor> cls_predictor;
 
-  RunRecModel(boxes, srcimg, rec_predictor, rec_text, rec_text_score,
-              charactor_dict, cls_predictor, use_direction_classify);
+  std::vector<double> time_info = {0, 0, 0};
+  for (int i = 0; i < cv_all_img_names.size(); ++i) {
+    std::cout << "The predict img: " << cv_all_img_names[i] << std::endl;
+    cv::Mat srcimg = cv::imread(cv_all_img_names[i], cv::IMREAD_COLOR);
 
-  auto end = std::chrono::system_clock::now();
-  auto duration =
-      std::chrono::duration_cast<std::chrono::microseconds>(end - start);
+    if (!srcimg.data) {
+      std::cerr << "[ERROR] image read failed! image path: " << cv_all_img_names[i] << std::endl;
+      exit(1);
+    }
 
-  //// visualization
-  auto img_vis = Visualization(srcimg, boxes);
+    int width = srcimg.cols;
+    int height = srcimg.rows;
+    std::vector<int> upper_left = {0, 0};
+    std::vector<int> upper_right = {width, 0};
+    std::vector<int> lower_right = {width, height};
+    std::vector<int> lower_left  = {0, height};
+    std::vector<std::vector<int>> box = {upper_left, upper_right, lower_right, lower_left};
+    std::vector<std::vector<std::vector<int>>> boxes = {box};
+
+    std::vector<std::string> rec_text;
+    std::vector<float> rec_text_score;
+    std::vector<double> times;
+    RunRecModel(boxes, srcimg, rec_predictor, rec_text, rec_text_score,
+                charactor_dict, cls_predictor, 0, &times);
+  
+    //// print recognized text
+    for (int i = 0; i < rec_text.size(); i++) {
+      std::cout << i << "\t" << rec_text[i] << "\t" << rec_text_score[i]
+                << std::endl;
+    }
+  }
+  // TODO: support autolog
+  if (strcmp(argv[9], "True") == 0) {
+    AutoLogger autolog(rec_model_file, 
+                       0,
+                       0,
+                       0,
+                       std::stoi(num_threads),
+                       std::stoi(batchsize), 
+                       "dynamic", 
+                       precision, 
+                       power_mode,
+                       time_info, 
+                       cv_all_img_names.size());
+    autolog.report();
+  }
+}
+
+int main(int argc, char **argv) {
+  check_params(argc, argv);
+  std::cout << "mode: " << argv[1] << endl;
 
-  //// print recognized text
-  for (int i = 0; i < rec_text.size(); i++) {
-    std::cout << i << "\t" << rec_text[i] << "\t" << rec_text_score[i]
-              << std::endl;
+  if (strcmp(argv[1], "system") == 0) {
+    system(argv);
   }
 
-  std::cout << "花费了"
-            << double(duration.count()) *
-                   std::chrono::microseconds::period::num /
-                   std::chrono::microseconds::period::den
-            << "秒" << std::endl;
+  if (strcmp(argv[1], "det") == 0) {
+    det(argc, argv);
+  }
+
+  if (strcmp(argv[1], "rec") == 0) {
+    rec(argc, argv);
+  }
 
   return 0;
-}
+}
\ No newline at end of file

doc/doc_ch/enhanced_ctc_loss.md

@@ -64,7 +64,7 @@ C-CTC Loss是CTC Loss + Center Loss的简称。 其中Center Loss出自论文 <
 
 以配置文件`configs/rec/ch_PP-OCRv2/ch_PP-OCRv2_rec.yml`为例， center提取命令如下所示:
 ```
-python tools/export_center.py -c configs/rec/ch_PP-OCRv2/ch_PP-OCRv2_rec.yml -o  Global.pretrained_model: "./output/rec_mobile_pp-OCRv2/best_accuracy"
+python tools/export_center.py -c configs/rec/ch_PP-OCRv2/ch_PP-OCRv2_rec.yml -o Global.pretrained_model="./output/rec_mobile_pp-OCRv2/best_accuracy"
 ```
 运行完后，会在PaddleOCR主目录下生成`train_center.pkl`.
 

ppocr/losses/rec_nrtr_loss.py

@@ -22,7 +22,7 @@ class NRTRLoss(nn.Layer):
             log_prb = F.log_softmax(pred, axis=1)
             non_pad_mask = paddle.not_equal(
                 tgt, paddle.zeros(
-                    tgt.shape, dtype='int64'))
+                    tgt.shape, dtype=tgt.dtype))
             loss = -(one_hot * log_prb).sum(axis=1)
             loss = loss.masked_select(non_pad_mask).mean()
         else:

ppocr/postprocess/__init__.py

@@ -29,10 +29,7 @@ from .rec_postprocess import CTCLabelDecode, AttnLabelDecode, SRNLabelDecode, Di
     TableLabelDecode, NRTRLabelDecode, SARLabelDecode , SEEDLabelDecode
 from .cls_postprocess import ClsPostProcess
 from .pg_postprocess import PGPostProcess
-
-if platform.system() != "Windows":
-    # pse is not support in Windows
-    from .pse_postprocess import PSEPostProcess
+from .pse_postprocess import PSEPostProcess
 
 
 def build_post_process(config, global_config=None):

ppocr/postprocess/pse_postprocess/pse/__init__.py

@@ -17,7 +17,12 @@ import subprocess
 
 python_path = sys.executable
 
-if subprocess.call('cd ppocr/postprocess/pse_postprocess/pse;{} setup.py build_ext --inplace;cd -'.format(python_path), shell=True) != 0:
-    raise RuntimeError('Cannot compile pse: {}'.format(os.path.dirname(os.path.realpath(__file__))))
+ori_path = os.getcwd()
+os.chdir('ppocr/postprocess/pse_postprocess/pse')
+if subprocess.call(
+        '{} setup.py build_ext --inplace'.format(python_path), shell=True) != 0:
+    raise RuntimeError('Cannot compile pse: {}'.format(
+        os.path.dirname(os.path.realpath(__file__))))
+os.chdir(ori_path)
 
-from .pse import pse
+from .pse import pse
\ No newline at end of file

PTDN/compare_results.py → test_tipc/compare_results.py

@@ -32,6 +32,7 @@ def run_shell_command(cmd):
     else:
         return None
 
+
 def parser_results_from_log_by_name(log_path, names_list):
     if not os.path.exists(log_path):
         raise ValueError("The log file {} does not exists!".format(log_path))
@@ -52,6 +53,7 @@ def parser_results_from_log_by_name(log_path, names_list):
         parser_results[name] = result
     return parser_results
 
+
 def load_gt_from_file(gt_file):
     if not os.path.exists(gt_file):
         raise ValueError("The log file {} does not exists!".format(gt_file))

PTDN/configs/ppocr_det_mobile_params.txt → test_tipc/configs/ppocr_det_mobile_params.txt

 ===========================train_params===========================
 model_name:ocr_det
 python:python3.7
-gpu_list:0|0,1
-Global.use_gpu:True|True
-Global.auto_cast:null
-Global.epoch_num:lite_train_infer=1|whole_train_infer=300
+gpu_list:0|0,1|10.21.226.181,10.21.226.133;0,1
+Global.use_gpu:True|True|True
+Global.auto_cast:fp32|amp
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=300
 Global.save_model_dir:./output/
-Train.loader.batch_size_per_card:lite_train_infer=2|whole_train_infer=4
+Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
 Global.pretrained_model:null
 train_model_name:latest
 train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
 null:null
 ##
 trainer:norm_train|pact_train|fpgm_train
-norm_train:tools/train.py -c tests/configs/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/MobileNetV3_large_x0_5_pretrained
-pact_train:deploy/slim/quantization/quant.py -c tests/configs/det_mv3_db.yml -o
-fpgm_train:deploy/slim/prune/sensitivity_anal.py -c tests/configs/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/det_mv3_db_v2.0_train/best_accuracy
+norm_train:tools/train.py -c test_tipc/configs/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/MobileNetV3_large_x0_5_pretrained
+pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/det_mv3_db.yml -o
+fpgm_train:deploy/slim/prune/sensitivity_anal.py -c test_tipc/configs/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/det_mv3_db_v2.0_train/best_accuracy
 distill_train:null
 null:null
 null:null
@@ -27,13 +27,13 @@ null:null
 ===========================infer_params===========================
 Global.save_inference_dir:./output/
 Global.pretrained_model:
-norm_export:tools/export_model.py -c tests/configs/det_mv3_db.yml -o 
-quant_export:deploy/slim/quantization/export_model.py -c tests/configs/det_mv3_db.yml -o 
-fpgm_export:deploy/slim/prune/export_prune_model.py -c tests/configs/det_mv3_db.yml -o 
+norm_export:tools/export_model.py -c test_tipc/configs/det_mv3_db.yml -o 
+quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/det_mv3_db.yml -o 
+fpgm_export:deploy/slim/prune/export_prune_model.py -c test_tipc/configs/det_mv3_db.yml -o 
 distill_export:null
 export1:null
 export2:null
-##
+inference_dir:null
 train_model:./inference/ch_ppocr_mobile_v2.0_det_train/best_accuracy
 infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
 infer_quant:False
@@ -98,3 +98,13 @@ null:null
 --benchmark:True
 null:null
 null:null
+===========================lite_params===========================
+inference:./ocr_db_crnn det
+infer_model:./models/ch_ppocr_mobile_v2.0_det_opt.nb|./models/ch_ppocr_mobile_v2.0_det_slim_opt.nb
+--cpu_threads:1|4
+--batch_size:1
+--power_mode:LITE_POWER_HIGH|LITE_POWER_LOW
+--image_dir:./test_data/icdar2015_lite/text_localization/ch4_test_images/|./test_data/icdar2015_lite/text_localization/ch4_test_images/img_233.jpg
+--config_dir:./config.txt
+--rec_dict_dir:./ppocr_keys_v1.txt
+--benchmark:True

test_tipc/configs/ppocrv2_det_mobile_params.txt

+===========================train_params===========================
+model_name:PPOCRv2_ocr_det
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_infer=1|whole_train_infer=500
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_infer=2|whole_train_infer=4
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
+null:null
+##
+trainer:norm_train|pact_train
+norm_train:tools/train.py -c configs/det/ch_PP-OCRv2/ch_PP-OCR_det_cml.yml -o 
+pact_train:deploy/slim/quantization/quant.py -c configs/det/ch_PP-OCRv2/ch_PP-OCR_det_cml.yml -o
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params=========================== 
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.pretrained_model:
+norm_export:tools/export_model.py -c configs/det/ch_PP-OCRv2/ch_PP-OCR_det_cml.yml -o 
+quant_export:deploy/slim/quantization/export_model.py -c configs/det/ch_PP-OCRv2/ch_PP-OCR_det_cml.yml -o 
+fpgm_export: 
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv2_det_infer/
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_det.py
+--use_gpu:True|False
+--enable_mkldnn:True|False
+--cpu_threads:1|6
+--rec_batch_num:1
+--use_tensorrt:False|True
+--precision:fp32|fp16|int8
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+null:null