From 6fcc2e715059ee99ca580d19b505800f1d746c67 Mon Sep 17 00:00:00 2001 From: WenmuZhou <572459439@qq.com> Date: Fri, 11 Mar 2022 04:04:11 +0000 Subject: [PATCH] fix bug in mem --- deploy/cpp_infer/src/ocr_rec.cpp | 221 ++++++++++++++++--------------- 1 file changed, 111 insertions(+), 110 deletions(-) diff --git a/deploy/cpp_infer/src/ocr_rec.cpp b/deploy/cpp_infer/src/ocr_rec.cpp index f1a97a99..25224f88 100644 --- a/deploy/cpp_infer/src/ocr_rec.cpp +++ b/deploy/cpp_infer/src/ocr_rec.cpp @@ -15,108 +15,115 @@ #include namespace PaddleOCR { - -void CRNNRecognizer::Run(std::vector img_list, std::vector *times) { - std::chrono::duration preprocess_diff = std::chrono::steady_clock::now() - std::chrono::steady_clock::now(); - std::chrono::duration inference_diff = std::chrono::steady_clock::now() - std::chrono::steady_clock::now(); - std::chrono::duration postprocess_diff = std::chrono::steady_clock::now() - std::chrono::steady_clock::now(); - - int img_num = img_list.size(); - std::vector width_list; - for (int i = 0; i < img_num; i++) { - width_list.push_back(float(img_list[i].cols) / img_list[i].rows); + +void CRNNRecognizer::Run(std::vector img_list, + std::vector *times) { + std::chrono::duration preprocess_diff = + std::chrono::steady_clock::now() - std::chrono::steady_clock::now(); + std::chrono::duration inference_diff = + std::chrono::steady_clock::now() - std::chrono::steady_clock::now(); + std::chrono::duration postprocess_diff = + std::chrono::steady_clock::now() - std::chrono::steady_clock::now(); + + int img_num = img_list.size(); + std::vector width_list; + for (int i = 0; i < img_num; i++) { + width_list.push_back(float(img_list[i].cols) / img_list[i].rows); + } + std::vector indices = Utility::argsort(width_list); + + for (int beg_img_no = 0; beg_img_no < img_num; + beg_img_no += this->rec_batch_num_) { + auto preprocess_start = std::chrono::steady_clock::now(); + int end_img_no = min(img_num, beg_img_no + this->rec_batch_num_); + float max_wh_ratio = 0; + for (int ino = beg_img_no; ino < end_img_no; ino++) { + int h = img_list[indices[ino]].rows; + int w = img_list[indices[ino]].cols; + float wh_ratio = w * 1.0 / h; + max_wh_ratio = max(max_wh_ratio, wh_ratio); } - std::vector indices = Utility::argsort(width_list); - - for (int beg_img_no = 0; beg_img_no < img_num; beg_img_no += this->rec_batch_num_) { - auto preprocess_start = std::chrono::steady_clock::now(); - int end_img_no = min(img_num, beg_img_no + this->rec_batch_num_); - float max_wh_ratio = 0; - for (int ino = beg_img_no; ino < end_img_no; ino ++) { - int h = img_list[indices[ino]].rows; - int w = img_list[indices[ino]].cols; - float wh_ratio = w * 1.0 / h; - max_wh_ratio = max(max_wh_ratio, wh_ratio); - } - std::vector norm_img_batch; - for (int ino = beg_img_no; ino < end_img_no; ino ++) { - cv::Mat srcimg; - img_list[indices[ino]].copyTo(srcimg); - cv::Mat resize_img; - this->resize_op_.Run(srcimg, resize_img, max_wh_ratio, this->use_tensorrt_); - this->normalize_op_.Run(&resize_img, this->mean_, this->scale_, this->is_scale_); - norm_img_batch.push_back(resize_img); - } - - int batch_width = int(ceilf(32 * max_wh_ratio)) - 1; - std::vector input(this->rec_batch_num_ * 3 * 32 * batch_width, 0.0f); - this->permute_op_.Run(norm_img_batch, input.data()); - auto preprocess_end = std::chrono::steady_clock::now(); - preprocess_diff += preprocess_end - preprocess_start; - - // Inference. - auto input_names = this->predictor_->GetInputNames(); - auto input_t = this->predictor_->GetInputHandle(input_names[0]); - input_t->Reshape({this->rec_batch_num_, 3, 32, batch_width}); - auto inference_start = std::chrono::steady_clock::now(); - input_t->CopyFromCpu(input.data()); - this->predictor_->Run(); - - std::vector predict_batch; - auto output_names = this->predictor_->GetOutputNames(); - auto output_t = this->predictor_->GetOutputHandle(output_names[0]); - auto predict_shape = output_t->shape(); - - int out_num = std::accumulate(predict_shape.begin(), predict_shape.end(), 1, - std::multiplies()); - predict_batch.resize(out_num); - - output_t->CopyToCpu(predict_batch.data()); - auto inference_end = std::chrono::steady_clock::now(); - inference_diff += inference_end - inference_start; - - // ctc decode - auto postprocess_start = std::chrono::steady_clock::now(); - for (int m = 0; m < predict_shape[0]; m++) { - std::vector str_res; - int argmax_idx; - int last_index = 0; - float score = 0.f; - int count = 0; - float max_value = 0.0f; - - for (int n = 0; n < predict_shape[1]; n++) { - argmax_idx = - int(Utility::argmax(&predict_batch[(m * predict_shape[1] + n) * predict_shape[2]], - &predict_batch[(m * predict_shape[1] + n + 1) * predict_shape[2]])); - max_value = - float(*std::max_element(&predict_batch[(m * predict_shape[1] + n) * predict_shape[2]], - &predict_batch[(m * predict_shape[1] + n + 1) * predict_shape[2]])); - - if (argmax_idx > 0 && (!(n > 0 && argmax_idx == last_index))) { - score += max_value; - count += 1; - str_res.push_back(label_list_[argmax_idx]); - } - last_index = argmax_idx; - } - score /= count; - if (isnan(score)) - continue; - for (int i = 0; i < str_res.size(); i++) { - std::cout << str_res[i]; - } - std::cout << "\tscore: " << score << std::endl; + int batch_width = 0; + std::vector norm_img_batch; + for (int ino = beg_img_no; ino < end_img_no; ino++) { + cv::Mat srcimg; + img_list[indices[ino]].copyTo(srcimg); + cv::Mat resize_img; + this->resize_op_.Run(srcimg, resize_img, max_wh_ratio, + this->use_tensorrt_); + this->normalize_op_.Run(&resize_img, this->mean_, this->scale_, + this->is_scale_); + norm_img_batch.push_back(resize_img); + batch_width = max(resize_img.cols, batch_width); + } + + std::vector input(this->rec_batch_num_ * 3 * 32 * batch_width, 0.0f); + this->permute_op_.Run(norm_img_batch, input.data()); + auto preprocess_end = std::chrono::steady_clock::now(); + preprocess_diff += preprocess_end - preprocess_start; + + // Inference. + auto input_names = this->predictor_->GetInputNames(); + auto input_t = this->predictor_->GetInputHandle(input_names[0]); + input_t->Reshape({this->rec_batch_num_, 3, 32, batch_width}); + auto inference_start = std::chrono::steady_clock::now(); + input_t->CopyFromCpu(input.data()); + this->predictor_->Run(); + + std::vector predict_batch; + auto output_names = this->predictor_->GetOutputNames(); + auto output_t = this->predictor_->GetOutputHandle(output_names[0]); + auto predict_shape = output_t->shape(); + + int out_num = std::accumulate(predict_shape.begin(), predict_shape.end(), 1, + std::multiplies()); + predict_batch.resize(out_num); + + output_t->CopyToCpu(predict_batch.data()); + auto inference_end = std::chrono::steady_clock::now(); + inference_diff += inference_end - inference_start; + + // ctc decode + auto postprocess_start = std::chrono::steady_clock::now(); + for (int m = 0; m < predict_shape[0]; m++) { + std::vector str_res; + int argmax_idx; + int last_index = 0; + float score = 0.f; + int count = 0; + float max_value = 0.0f; + + for (int n = 0; n < predict_shape[1]; n++) { + argmax_idx = int(Utility::argmax( + &predict_batch[(m * predict_shape[1] + n) * predict_shape[2]], + &predict_batch[(m * predict_shape[1] + n + 1) * predict_shape[2]])); + max_value = float(*std::max_element( + &predict_batch[(m * predict_shape[1] + n) * predict_shape[2]], + &predict_batch[(m * predict_shape[1] + n + 1) * predict_shape[2]])); + + if (argmax_idx > 0 && (!(n > 0 && argmax_idx == last_index))) { + score += max_value; + count += 1; + str_res.push_back(label_list_[argmax_idx]); } - auto postprocess_end = std::chrono::steady_clock::now(); - postprocess_diff += postprocess_end - postprocess_start; + last_index = argmax_idx; + } + score /= count; + if (isnan(score)) + continue; + for (int i = 0; i < str_res.size(); i++) { + std::cout << str_res[i]; + } + std::cout << "\tscore: " << score << std::endl; } - times->push_back(double(preprocess_diff.count() * 1000)); - times->push_back(double(inference_diff.count() * 1000)); - times->push_back(double(postprocess_diff.count() * 1000)); + auto postprocess_end = std::chrono::steady_clock::now(); + postprocess_diff += postprocess_end - postprocess_start; + } + times->push_back(double(preprocess_diff.count() * 1000)); + times->push_back(double(inference_diff.count() * 1000)); + times->push_back(double(postprocess_diff.count() * 1000)); } - void CRNNRecognizer::LoadModel(const std::string &model_dir) { // AnalysisConfig config; paddle_infer::Config config; @@ -130,23 +137,17 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) { if (this->precision_ == "fp16") { precision = paddle_infer::Config::Precision::kHalf; } - if (this->precision_ == "int8") { + if (this->precision_ == "int8") { precision = paddle_infer::Config::Precision::kInt8; - } - config.EnableTensorRtEngine( - 1 << 20, 10, 3, - precision, - false, false); + } + config.EnableTensorRtEngine(1 << 20, 10, 3, precision, false, false); std::map> min_input_shape = { - {"x", {1, 3, 32, 10}}, - {"lstm_0.tmp_0", {10, 1, 96}}}; + {"x", {1, 3, 32, 10}}, {"lstm_0.tmp_0", {10, 1, 96}}}; std::map> max_input_shape = { - {"x", {1, 3, 32, 2000}}, - {"lstm_0.tmp_0", {1000, 1, 96}}}; + {"x", {1, 3, 32, 2000}}, {"lstm_0.tmp_0", {1000, 1, 96}}}; std::map> opt_input_shape = { - {"x", {1, 3, 32, 320}}, - {"lstm_0.tmp_0", {25, 1, 96}}}; + {"x", {1, 3, 32, 320}}, {"lstm_0.tmp_0", {25, 1, 96}}}; config.SetTRTDynamicShapeInfo(min_input_shape, max_input_shape, opt_input_shape); @@ -168,7 +169,7 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) { config.SwitchIrOptim(true); config.EnableMemoryOptim(); -// config.DisableGlogInfo(); + // config.DisableGlogInfo(); this->predictor_ = CreatePredictor(config); } -- GitLab