diff --git a/research/cv/DeepID/README.md b/research/cv/DeepID/README.md
index faea081b4591e8d41717d8b2e444880bf3ef7724..8647d1a34dcc67c3250f04874377b836c5cb0256 100644
--- a/research/cv/DeepID/README.md
+++ b/research/cv/DeepID/README.md
@@ -54,7 +54,7 @@ python crop.py
 
 After cropping images, change **src_folder** in split.py and split data into two set, One is for train and one is for valid.
 
-```python
+```bash
 python split.py
 ```
 
@@ -75,11 +75,13 @@ python split.py
 ```shell
 .
 鈹斺攢 DeepID
+  鈹溾攢鈹€ ascend310_infer                 // 310 infer directory
   鈹溾攢 README.md                        // Descriptions about DeepID
   鈹溾攢 scripts
     鈹溾攢 run_standalone_train_ascend.sh // Train standalone
     鈹溾攢 run_distribute_train_ascend.sh // Train distribute
     鈹斺攢 run_eval_ascend.sh             // Evaluation
+    鈹斺攢 run_infer_310.sh               // 310 inference
   鈹溾攢src
     鈹溾攢 dataset.py                     // Prepare dataset
     鈹溾攢 loss.py                        // Loss function
@@ -92,11 +94,13 @@ python split.py
   鈹溾攢 eval.py                          // Evaluation script
   鈹溾攢 train.py                         // Train script
   鈹溾攢 export.py                        // Export mindir script
+  鈹溾攢 preprocess.py                    // Convert images and labels to bin
+  鈹溾攢 postprocess.py                   // Calculate accuracy
 ```
 
 ## Script parameters
 
-```shell
+```bash
 'data_url':'./data/'      # Dataset path
 'epochs':200              # Total epochs
 'lr':1e-4                 # Learning rate
@@ -132,6 +136,25 @@ sh run_distribute_train.sh [DEVICE_NUM] [DISTRIBUTE] [RANK_TABLE_FILE]
 sh eval_ascend.sh [DEVICE_NUM] [DEVICE_ID]
 ```
 
+## [Ascen 310 infer](#contents)
+
+### Export MindIR
+
+```bash
+python export.py
+```
+
+### Infer on Ascend 310
+
+```bash
+cd scripts
+bash run_infer_310.sh [MINDIR_PATH] [DATA_PATH] [DEVICE_ID]
+```
+
+- `MINDIR_PATH` Directionary of MINDIR
+- `DATA_PATH` Directionary of dataset
+- `DEVICE_ID` Optional, default 0
+
 # [Result](#Contents)
 
 The evaluation results will be saved in the sample path in a log file named "log_eval.txt". You can find results similar to the following in the log.
@@ -173,6 +196,16 @@ Valid dataset accuracy: 0.9683
 | batch_size          | 512                             |
 | outputs             | Accuracy = 96.83%               |
 
+| Parameters          | Ascend 310                      |
+| ------------------- | ---------------------------     |
+| Model Version       | DeepID                          |
+| Resource            | Ascend                          |
+| Uploaded Date       | 11/30/2021 (month/day/year)     |
+| MindSpore Version   | 1.3.1                           |
+| Dataset             | Youtube Face                    |
+| batch_size          | 1                               |
+| outputs             | Accuracy = 96.83%               |
+
 # [ModelZoo Homepage](#Contents)
 
 Please check the official [homepage](https://gitee.com/mindspore/models).
\ No newline at end of file
diff --git a/research/cv/DeepID/ascend310_infer/inc/utils.h b/research/cv/DeepID/ascend310_infer/inc/utils.h
new file mode 100644
index 0000000000000000000000000000000000000000..0b400632f51ee34707a5becc00f7f5ba05899b3a
--- /dev/null
+++ b/research/cv/DeepID/ascend310_infer/inc/utils.h
@@ -0,0 +1,33 @@
+/**
+ * Copyright 2021 Huawei Technologies Co., Ltd
+ * 
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ * 
+ * http://www.apache.org/licenses/LICENSE-2.0
+ * 
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef MINDSPORE_INFERENCE_UTILS_H_
+#define MINDSPORE_INFERENCE_UTILS_H_
+
+#include <sys/stat.h>
+#include <dirent.h>
+#include <vector>
+#include <string>
+#include <memory>
+#include "include/api/types.h"
+
+DIR *OpenDir(std::string_view dirName);
+std::string RealPath(std::string_view path);
+mindspore::MSTensor ReadFileToTensor(const std::string &file);
+int WriteResult(const std::string& imageFile, const std::vector<mindspore::MSTensor> &outputs);
+std::vector<std::string> GetAllFiles(std::string dir_name);
+
+#endif
diff --git a/research/cv/DeepID/ascend310_infer/src/CMakeLists.txt b/research/cv/DeepID/ascend310_infer/src/CMakeLists.txt
new file mode 100644
index 0000000000000000000000000000000000000000..14e676821a4936c03e98b9299b3b5f5e4496a8ea
--- /dev/null
+++ b/research/cv/DeepID/ascend310_infer/src/CMakeLists.txt
@@ -0,0 +1,14 @@
+cmake_minimum_required(VERSION 3.14.1)
+project(MindSporeCxxTestcase[CXX])
+add_compile_definitions(_GLIBCXX_USE_CXX11_ABI=0)
+set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -O0 -g -std=c++17 -Werror -Wall -fPIE -Wl,--allow-shlib-undefined")
+set(PROJECT_SRC_ROOT ${CMAKE_CURRENT_LIST_DIR}/)
+option(MINDSPORE_PATH "mindspore install path" "")
+include_directories(${MINDSPORE_PATH})
+include_directories(${MINDSPORE_PATH}/include)
+include_directories(${PROJECT_SRC_ROOT}/../)
+find_library(MS_LIB libmindspore.so ${MINDSPORE_PATH}/lib)
+file(GLOB_RECURSE MD_LIB ${MINDSPORE_PATH}/_c_dataengine*)
+find_package(gflags REQUIRED)
+add_executable(main main.cc utils.cc)
+target_link_libraries(main ${MS_LIB} ${MD_LIB} gflags)
diff --git a/research/cv/DeepID/ascend310_infer/src/build.sh b/research/cv/DeepID/ascend310_infer/src/build.sh
new file mode 100644
index 0000000000000000000000000000000000000000..f1945fe52b0924fe9826042de2f1b7c7f02ebe98
--- /dev/null
+++ b/research/cv/DeepID/ascend310_infer/src/build.sh
@@ -0,0 +1,18 @@
+#!/bin/bash
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+
+cmake . -DMINDSPORE_PATH="`pip3 show mindspore-ascend | grep Location | awk '{print $2"/mindspore"}' | xargs realpath`"
+make
diff --git a/research/cv/DeepID/ascend310_infer/src/main.cc b/research/cv/DeepID/ascend310_infer/src/main.cc
new file mode 100644
index 0000000000000000000000000000000000000000..a5ac8bc96188f6c91c0a0daf52889d888800d4e7
--- /dev/null
+++ b/research/cv/DeepID/ascend310_infer/src/main.cc
@@ -0,0 +1,139 @@
+/**
+ * Copyright 2021 Huawei Technologies Co., Ltd
+ * 
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ * 
+ * http://www.apache.org/licenses/LICENSE-2.0
+ * 
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+#include <sys/time.h>
+#include <gflags/gflags.h>
+#include <dirent.h>
+#include <iostream>
+#include <string>
+#include <algorithm>
+#include <iosfwd>
+#include <vector>
+#include <fstream>
+#include <sstream>
+
+#include "../inc/utils.h"
+#include "include/dataset/execute.h"
+#include "include/dataset/transforms.h"
+#include "include/dataset/vision.h"
+#include "include/dataset/vision_ascend.h"
+#include "include/api/types.h"
+#include "include/api/model.h"
+#include "include/api/serialization.h"
+#include "include/api/context.h"
+using mindspore::dataset::vision::Decode;
+using mindspore::dataset::vision::Resize;
+using mindspore::dataset::vision::CenterCrop;
+using mindspore::dataset::vision::Normalize;
+using mindspore::dataset::vision::HWC2CHW;
+using mindspore::dataset::TensorTransform;
+using mindspore::Context;
+using mindspore::Serialization;
+using mindspore::Model;
+using mindspore::Status;
+using mindspore::ModelType;
+using mindspore::GraphCell;
+using mindspore::kSuccess;
+using mindspore::MSTensor;
+using mindspore::dataset::Execute;
+
+DEFINE_string(mindir_path, "", "mindir path");
+DEFINE_string(img_path, ".", "img path");
+DEFINE_int32(device_id, 0, "device id");
+
+int main(int argc, char **argv) {
+    gflags::ParseCommandLineFlags(&argc, &argv, true);
+    if (RealPath(FLAGS_mindir_path).empty()) {
+        std::cout << "Invalid mindir" << std::endl;
+        return 1;
+    }
+    if (RealPath(FLAGS_img_path).empty()) {
+        std::cout << "Invalid img" << std::endl;
+        return 1;
+    }
+
+    auto context = std::make_shared<Context>();
+    auto ascend310 = std::make_shared<mindspore::Ascend310DeviceInfo>();
+    ascend310->SetDeviceID(FLAGS_device_id);
+    context->MutableDeviceInfo().push_back(ascend310);
+    mindspore::Graph graph;
+    Serialization::Load(FLAGS_mindir_path, ModelType::kMindIR, &graph);
+    Model model;
+    Status ret = model.Build(GraphCell(graph), context);
+    if (ret != kSuccess) {
+        std::cout << "ERROR: Build failed." << std::endl;
+        return 1;
+    }
+
+    std::vector<MSTensor> modelInputs = model.GetInputs();
+    auto all_img_files = GetAllFiles(FLAGS_img_path);
+    if (all_img_files.empty()) {
+        std::cout << "ERROR: no input img." << std::endl;
+        return 1;
+    }
+
+    std::map<double, double> costTime_map;
+    size_t size = all_img_files.size();
+
+    for (size_t i = 0; i < size; ++i) {
+        struct timeval start = {0};
+        struct timeval end = {0};
+        double startTimeMs;
+        double endTimeMs;
+
+        std::vector<MSTensor> inputs;
+        std::vector<MSTensor> outputs;
+        std::cout << "Start predict input files:" << all_img_files[i] <<std::endl;
+
+        auto image = ReadFileToTensor(all_img_files[i]);
+
+        inputs.emplace_back(modelInputs[0].Name(), modelInputs[0].DataType(), modelInputs[0].Shape(),
+                            image.Data().get(), image.DataSize());
+
+        gettimeofday(&start, nullptr);
+        ret = model.Predict(inputs, &outputs);
+        gettimeofday(&end, nullptr);
+        if (ret != kSuccess) {
+            std::cout << "Predict " << all_img_files[i] << " failed." << std::endl;
+            return 1;
+        }
+        std::cout << "Predict Successful!" << std::endl;
+        startTimeMs = (1.0 * start.tv_sec * 1000000 + start.tv_usec) / 1000;
+        endTimeMs = (1.0 * end.tv_sec * 1000000 + end.tv_usec) / 1000;
+        costTime_map.insert(std::pair<double, double>(startTimeMs, endTimeMs));
+        int rst = WriteResult(all_img_files[i], outputs);
+        if (rst != 0) {
+            std::cout << "write result failed." << std::endl;
+            return rst;
+        }
+    }
+    double average = 0.0;
+    int inferCount = 0;
+
+    for (auto iter = costTime_map.begin(); iter != costTime_map.end(); iter++) {
+        average += iter->second - iter->first;
+        inferCount++;
+    }
+    average = average / inferCount;
+    std::stringstream timeCost;
+    timeCost << "NN inference cost average time: "<< average << " ms of infer_count " << inferCount << std::endl;
+    std::cout << "NN inference cost average time: "<< average << "ms of infer_count " << inferCount << std::endl;
+    std::string fileName = "./time_Result" + std::string("/test_perform_static.txt");
+    std::ofstream fileStream(fileName.c_str(), std::ios::trunc);
+    fileStream << timeCost.str();
+    fileStream.close();
+    costTime_map.clear();
+    return 0;
+}
diff --git a/research/cv/DeepID/ascend310_infer/src/utils.cc b/research/cv/DeepID/ascend310_infer/src/utils.cc
new file mode 100644
index 0000000000000000000000000000000000000000..d545acd312fbd7dc3ae2da3a7d29aea3c6db86e3
--- /dev/null
+++ b/research/cv/DeepID/ascend310_infer/src/utils.cc
@@ -0,0 +1,197 @@
+/**
+ * Copyright 2021 Huawei Technologies Co., Ltd
+ * 
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ * 
+ * http://www.apache.org/licenses/LICENSE-2.0
+ * 
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include <fstream>
+#include <algorithm>
+#include <iostream>
+#include "inc/utils.h"
+
+using mindspore::MSTensor;
+using mindspore::DataType;
+
+std::vector<std::vector<std::string>> GetAllInputData(std::string dir_name) {
+  std::vector<std::vector<std::string>> ret;
+
+  DIR *dir = OpenDir(dir_name);
+  if (dir == nullptr) {
+    return {};
+  }
+  struct dirent *filename;
+  /* read all the files in the dir ~ */
+  std::vector<std::string> sub_dirs;
+  while ((filename = readdir(dir)) != nullptr) {
+    std::string d_name = std::string(filename->d_name);
+    // get rid of "." and ".."
+    if (d_name == "." || d_name == ".." || d_name.empty()) {
+      continue;
+    }
+    std::string dir_path = RealPath(std::string(dir_name) + "/" + filename->d_name);
+    struct stat s;
+    lstat(dir_path.c_str(), &s);
+    if (!S_ISDIR(s.st_mode)) {
+      continue;
+    }
+
+    sub_dirs.emplace_back(dir_path);
+  }
+  std::sort(sub_dirs.begin(), sub_dirs.end());
+
+  (void)std::transform(sub_dirs.begin(), sub_dirs.end(), std::back_inserter(ret),
+                       [](const std::string &d) { return GetAllFiles(d); });
+
+  return ret;
+}
+
+
+std::vector<std::string> GetAllFiles(std::string dir_name) {
+  struct dirent *filename;
+  DIR *dir = OpenDir(dir_name);
+  if (dir == nullptr) {
+    return {};
+  }
+
+  std::vector<std::string> res;
+  while ((filename = readdir(dir)) != nullptr) {
+    std::string d_name = std::string(filename->d_name);
+    if (d_name == "." || d_name == ".." || d_name.size() <= 3) {
+      continue;
+    }
+    res.emplace_back(std::string(dir_name) + "/" + filename->d_name);
+  }
+  std::sort(res.begin(), res.end());
+
+  return res;
+}
+
+
+std::vector<std::string> GetAllFiles(std::string_view dirName) {
+  struct dirent *filename;
+  DIR *dir = OpenDir(dirName);
+  if (dir == nullptr) {
+    return {};
+  }
+  std::vector<std::string> res;
+  while ((filename = readdir(dir)) != nullptr) {
+    std::string dName = std::string(filename->d_name);
+    if (dName == "." || dName == ".." || filename->d_type != DT_REG) {
+      continue;
+    }
+    res.emplace_back(std::string(dirName) + "/" + filename->d_name);
+  }
+  std::sort(res.begin(), res.end());
+  for (auto &f : res) {
+    std::cout << "image file: " << f << std::endl;
+  }
+  return res;
+}
+
+
+int WriteResult(const std::string& imageFile, const std::vector<MSTensor> &outputs) {
+  std::string homePath = "./result_Files";
+  const int INVALID_POINTER = -1;
+  const int ERROR = -2;
+  for (size_t i = 0; i < outputs.size(); ++i) {
+    size_t outputSize;
+    std::shared_ptr<const void> netOutput;
+    netOutput = outputs[i].Data();
+    outputSize = outputs[i].DataSize();
+    int pos = imageFile.rfind('/');
+    std::string fileName(imageFile, pos + 1);
+    fileName.replace(fileName.find('.'), fileName.size() - fileName.find('.'), '_' + std::to_string(i) + ".bin");
+    std::string outFileName = homePath + "/" + fileName;
+    FILE *outputFile = fopen(outFileName.c_str(), "wb");
+    if (outputFile == nullptr) {
+        std::cout << "open result file " << outFileName << " failed" << std::endl;
+        return INVALID_POINTER;
+    }
+    size_t size = fwrite(netOutput.get(), sizeof(char), outputSize, outputFile);
+    if (size != outputSize) {
+        fclose(outputFile);
+        outputFile = nullptr;
+        std::cout << "write result file " << outFileName << " failed, write size[" << size <<
+            "] is smaller than output size[" << outputSize << "], maybe the disk is full." << std::endl;
+        return ERROR;
+    }
+    fclose(outputFile);
+    outputFile = nullptr;
+  }
+  return 0;
+}
+
+mindspore::MSTensor ReadFileToTensor(const std::string &file) {
+  if (file.empty()) {
+    std::cout << "Pointer file is nullptr" << std::endl;
+    return mindspore::MSTensor();
+  }
+
+  std::ifstream ifs(file);
+  if (!ifs.good()) {
+    std::cout << "File: " << file << " is not exist" << std::endl;
+    return mindspore::MSTensor();
+  }
+
+  if (!ifs.is_open()) {
+    std::cout << "File: " << file << "open failed" << std::endl;
+    return mindspore::MSTensor();
+  }
+
+  ifs.seekg(0, std::ios::end);
+  size_t size = ifs.tellg();
+  mindspore::MSTensor buffer(file, mindspore::DataType::kNumberTypeUInt8, {static_cast<int64_t>(size)}, nullptr, size);
+
+  ifs.seekg(0, std::ios::beg);
+  ifs.read(reinterpret_cast<char *>(buffer.MutableData()), size);
+  ifs.close();
+
+  return buffer;
+}
+
+
+DIR *OpenDir(std::string_view dirName) {
+  if (dirName.empty()) {
+    std::cout << " dirName is null ! " << std::endl;
+    return nullptr;
+  }
+  std::string realPath = RealPath(dirName);
+  struct stat s;
+  lstat(realPath.c_str(), &s);
+  if (!S_ISDIR(s.st_mode)) {
+    std::cout << "dirName is not a valid directory !" << std::endl;
+    return nullptr;
+  }
+  DIR *dir;
+  dir = opendir(realPath.c_str());
+  if (dir == nullptr) {
+    std::cout << "Can not open dir " << dirName << std::endl;
+    return nullptr;
+  }
+  std::cout << "Successfully opened the dir " << dirName << std::endl;
+  return dir;
+}
+
+std::string RealPath(std::string_view path) {
+  char realPathMem[PATH_MAX] = {0};
+  char *realPathRet = nullptr;
+  realPathRet = realpath(path.data(), realPathMem);
+  if (realPathRet == nullptr) {
+    std::cout << "File: " << path << " is not exist.";
+    return "";
+  }
+
+  std::string realPath(realPathMem);
+  std::cout << path << " realpath is: " << realPath << std::endl;
+  return realPath;
+}
diff --git a/research/cv/DeepID/postprocess.py b/research/cv/DeepID/postprocess.py
new file mode 100644
index 0000000000000000000000000000000000000000..68c20f8e3ff7d99433832b8e53a7efe085e90b2e
--- /dev/null
+++ b/research/cv/DeepID/postprocess.py
@@ -0,0 +1,51 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+"""post process for 310 inference"""
+import os
+import argparse
+import numpy as np
+
+
+parser = argparse.ArgumentParser(description='PostProcess args')
+parser.add_argument('--result_path', type=str, required=True, help='Dataset path')
+parser.add_argument('--ori_path', type=str, required=True, help='Train output path')
+
+args_opt = parser.parse_args()
+
+
+if __name__ == '__main__':
+    result_path = args_opt.result_path
+    ori_path = args_opt.ori_path
+    count = 0
+
+    result_file = os.listdir(result_path)
+    ori_path_file = os.listdir(ori_path)
+
+    assert len(result_file) == len(ori_path_file)
+
+    total_num = len(result_file)
+
+    for i in range(total_num):
+        ori_label_name = os.path.join(ori_path, 'sop_' + str(i) + '.bin')
+        result_label = os.path.join(result_path, 'sop_' + str(i) + '_0.bin')
+        ori_label = np.fromfile(ori_label_name, np.int64)
+        result_label = np.argmax(np.fromfile(result_label, np.float32), axis=0)
+        print("Start processing", ori_label_name)
+        if ori_label == result_label:
+            count += 1
+
+    acc = 100 * count / total_num
+    print("=" * 20, "Convert bin files finished", "=" * 20)
+    print("Accuracy is", round(acc, 2), "%")
diff --git a/research/cv/DeepID/preprocess.py b/research/cv/DeepID/preprocess.py
new file mode 100644
index 0000000000000000000000000000000000000000..c9795af56b432bc6c7e17e1504d52e6442136218
--- /dev/null
+++ b/research/cv/DeepID/preprocess.py
@@ -0,0 +1,56 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+"""Preprocess for DeepID"""
+import os
+import time
+import argparse
+
+from src.dataset import dataloader
+
+parser = argparse.ArgumentParser(description='DeepID_preprocess')
+
+parser.add_argument('--data_url', type=str, default='data/', help='Dataset path')
+parser.add_argument('--save_url', type=str, default='../bin_data', help='Dataset path')
+parser.add_argument('--batch_size', type=int, default=1, help='Batch Size')
+parser.add_argument('--mode', type=str, default='valid', help='dataset mode')
+
+if __name__ == '__main__':
+    args_opt = parser.parse_args()
+
+    valid_dataset, valid_dataset_length = dataloader(args_opt.data_url, epoch=1,
+                                                     mode=args_opt.mode, batch_size=args_opt.batch_size)
+
+    valid_dataset_iter = valid_dataset.create_dict_iterator()
+    print('Valid dataset length:', valid_dataset_length)
+
+    img_path = os.path.join(args_opt.save_url, "img_data")
+    label_path = os.path.join(args_opt.save_url, "label")
+    if not os.path.exists(img_path):
+        os.makedirs(img_path)
+    if not os.path.exists(label_path):
+        os.makedirs(label_path)
+
+    for idx, data in enumerate(valid_dataset_iter):
+        step_begin_time = time.time()
+        img_valid = data['image']
+        label_valid = data['label']
+        file_name = "sop_" + str(idx) + ".bin"
+        img_file_path = os.path.join(img_path, file_name)
+        label_file_path = os.path.join(label_path, file_name)
+        img_valid.asnumpy().tofile(img_file_path)
+        label_valid.asnumpy().tofile(label_file_path)
+        print('Finish processing img', idx, "saving as", file_name)
+
+    print("=" * 20, "export bin files finished", "=" * 20)
diff --git a/research/cv/DeepID/scripts/run_infer_310.sh b/research/cv/DeepID/scripts/run_infer_310.sh
new file mode 100644
index 0000000000000000000000000000000000000000..f00242954bca67bc7b4e63fe47bd61d3be08e30e
--- /dev/null
+++ b/research/cv/DeepID/scripts/run_infer_310.sh
@@ -0,0 +1,120 @@
+#!/bin/bash
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+
+if [[ $# -lt 2 || $# -gt 3 ]]; then
+    echo "Usage: bash run_infer_310.sh [MINDIR_PATH] [DATA_PATH] [DEVICE_ID]
+    DEVICE_ID is optional, it can be set by environment variable device_id, otherwise the value is zero"
+exit 1
+fi
+
+get_real_path(){
+    if [ "${1:0:1}" == "/" ]; then
+        echo "$1"
+    else
+        echo "$(realpath -m $PWD/$1)"
+    fi
+}
+model=$(get_real_path $1)
+data_path=$(get_real_path $2)
+
+scripts_path=$PWD
+
+device_id=0
+if [ $# == 3 ]; then
+    device_id=$3
+fi
+
+echo "mindir name: "$model
+echo "dataset path: "$data_path
+echo "device id: "$device_id
+
+export ASCEND_HOME=/usr/local/Ascend/
+if [ -d ${ASCEND_HOME}/ascend-toolkit ]; then
+    export PATH=$ASCEND_HOME/fwkacllib/bin:$ASCEND_HOME/fwkacllib/ccec_compiler/bin:$ASCEND_HOME/ascend-toolkit/latest/fwkacllib/ccec_compiler/bin:$ASCEND_HOME/ascend-toolkit/latest/atc/bin:$PATH
+    export LD_LIBRARY_PATH=$ASCEND_HOME/fwkacllib/lib64:/usr/local/lib:$ASCEND_HOME/ascend-toolkit/latest/atc/lib64:$ASCEND_HOME/ascend-toolkit/latest/fwkacllib/lib64:$ASCEND_HOME/driver/lib64:$ASCEND_HOME/add-ons:$LD_LIBRARY_PATH
+    export TBE_IMPL_PATH=$ASCEND_HOME/ascend-toolkit/latest/opp/op_impl/built-in/ai_core/tbe
+    export PYTHONPATH=$ASCEND_HOME/fwkacllib/python/site-packages:${TBE_IMPL_PATH}:$ASCEND_HOME/ascend-toolkit/latest/fwkacllib/python/site-packages:$PYTHONPATH
+    export ASCEND_OPP_PATH=$ASCEND_HOME/ascend-toolkit/latest/opp
+else
+    export PATH=$ASCEND_HOME/fwkacllib/bin:$ASCEND_HOME/fwkacllib/ccec_compiler/bin:$ASCEND_HOME/atc/ccec_compiler/bin:$ASCEND_HOME/atc/bin:$PATH
+    export LD_LIBRARY_PATH=$ASCEND_HOME/fwkacllib/lib64:/usr/local/lib:$ASCEND_HOME/atc/lib64:$ASCEND_HOME/acllib/lib64:$ASCEND_HOME/driver/lib64:$ASCEND_HOME/add-ons:$LD_LIBRARY_PATH
+    export PYTHONPATH=$ASCEND_HOME/fwkacllib/python/site-packages:$ASCEND_HOME/atc/python/site-packages:$PYTHONPATH
+    export ASCEND_OPP_PATH=$ASCEND_HOME/opp
+fi
+
+
+function preprocess_data()
+{
+    cd ../ || exit
+    if [ -d bin_data ]; then
+        rm -rf ./bin_data
+    fi
+    mkdir bin_data
+    cd $scripts_path
+    python3.7 ../preprocess.py --data_url=$data_path &> preprocess.log
+}
+
+function compile_app()
+{   
+    cd ../ascend310_infer/src/ || exit
+    if [ -f "Makefile" ]; then
+        make clean
+    fi
+    bash build.sh &> build.log
+
+}
+
+function infer()
+{
+    cd - || exit
+    if [ -d result_Files ]; then
+        rm -rf ./result_Files
+    fi
+    if [ -d time_Result ]; then
+        rm -rf ./time_Result
+    fi
+    mkdir result_Files
+    mkdir time_Result
+
+    ../ascend310_infer/src/main --mindir_path=$model --img_path="../bin_data/img_data" --device_id=$device_id &> infer.log
+}
+
+function postprocess()
+{
+    cd $scripts_path
+    python3.7 ../postprocess.py --result_path='./result_Files' --ori_path='../bin_data/label' &> postprocess.log
+}
+
+preprocess_data
+if [ $? -ne 0 ]; then
+    echo "preprocess data failed"
+    exit 1
+fi
+compile_app
+if [ $? -ne 0 ]; then
+    echo "compile app code failed"
+    exit 1
+fi
+infer
+if [ $? -ne 0 ]; then
+    echo "execute inference failed"
+    exit 1
+fi
+postprocess
+if [ $? -ne 0 ]; then
+    echo "execute inference failed"
+    exit 1
+fi