README.md

@@ -17,9 +17,10 @@ M. Verucchi, L. Bartoli, F. Bagni, F. Gatti, P. Burgio and M. Bertogna, "Real-Ti
     - [2)Export weights for DLA34 and ResNet101](#2export-weights-for-dla34-and-resnet101)
     - [3)Export weights for CenterNet](#3export-weights-for-centernet)
     - [4)Export weights for MobileNetSSD](#4export-weights-for-mobilenetssd)
-  - [Run the demo](#run-the-demo)
+  - [How to convert weights](#how-to-convert-weights)
     - [FP16 inference](#fp16-inference)
     - [INT8 inference](#int8-inference)
+  - [Run the demo](#run-the-demo)
   - [mAP demo](#map-demo)
   - [Existing tests and supported networks](#existing-tests-and-supported-networks)
   - [References](#references)
@@ -113,26 +114,8 @@ cd pytorch-ssd
 conda env create -f env_mobv2ssd.yml
 python run_ssd_live_demo.py mb2-ssd-lite <pth-model-fil> <labels-file>
 ```
-## Run the demo
 
-To run the an object detection demo follow these steps (example with yolov3):
-```
-rm yolo3_fp32.rt        # be sure to delete(or move) old tensorRT files
-./test_yolo3            # run the yolo test (is slow)
-./demo yolo3_fp32.rt ../demo/yolo_test.mp4 y
-```
-In general the demo program takes 4 parameters:
-```
-./demo <network-rt-file> <path-to-video> <kind-of-network> <number-of-classes>
-```
-where
-*  ```<network-rt-file>``` is the rt file generated by a test
-*  ```<<path-to-video>``` is the path to a video file or a camera input  
-*  ```<kind-of-network>``` is the type of network. Thee types are currently supported: ```y``` (YOLO family), ```c``` (CenterNet family) and ```m``` (MobileNet-SSD family)
-*  ```<number-of-classes>```is the number of classes the network is trained on
-N.b. By default it is used FP32 inference
-
-![demo](https://user-images.githubusercontent.com/11562617/72547657-540e7800-388d-11ea-83c6-49dfea2a0607.gif)
+## How to convert weights
 
 ### FP16 inference
 
@@ -187,6 +170,29 @@ rm yolo3_fp32.rt                   # be sure to delete(or move) old tensorRT fil
 ./test_rtinference yolo3_fp32.rt 4 # test with a batch size of 4
 ```
 
+## Run the demo
+
+To run the an object detection demo follow these steps (example with yolov3):
+```
+rm yolo3_fp32.rt        # be sure to delete(or move) old tensorRT files
+./test_yolo3            # run the yolo test (is slow)
+./demo yolo3_fp32.rt ../demo/yolo_test.mp4 y # add parameter y for yolo. c for CNET & m for Mobilenet.
+./demo yolo3_fp32.rt ../demo/yolo_test.mp4 y 1 #if number of classes not 80. Add class number parameter.
+```
+In general the demo program takes 5 parameters:
+```
+./demo <network-rt-file> <path-to-video> <kind-of-network> <number-of-classes> <output-type>
+```
+where
+*  ```<network-rt-file>``` is the rt file generated by a test
+*  ```<<path-to-video>``` is the path to a video file or a camera input  
+*  ```<kind-of-network>``` is the type of network. Thee types are currently supported: ```y``` (YOLO family), ```c``` (CenterNet family) and ```m``` (MobileNet-SSD family)
+*  ```<number-of-classes>```is the number of classes the network is trained on
+N.b. By default it is used FP32 inference
+* ```<output-type>``` benchmark or save_result. Adding benchmark will not show opencv detection video allowing demo to be run from terminal, providing performance results without showing the video. Adding save_result will save output of detection to results.mp4.
+
+![demo](https://user-images.githubusercontent.com/11562617/72547657-540e7800-388d-11ea-83c6-49dfea2a0607.gif)
+
 ## mAP demo
 
 To compute mAP, precision, recall and f1score, run the map_demo.

demo/demo/demo.cpp

@@ -10,6 +10,7 @@
 
 bool gRun;
 bool SAVE_RESULT = false;
+bool BENCHMARK = false;
 
 void sig_handler(int signo) {
     std::cout<<"request gateway stop\n";
@@ -33,7 +34,13 @@ int main(int argc, char *argv[]) {
         ntype = argv[3][0]; 
     int n_classes = 80;
     if(argc > 4)
-        n_classes = atoi(argv[4]); 
+        n_classes = atoi(argv[4]);
+    if(argc > 5 && strcmp(argv[5], "benchmark") == 0) {
+        BENCHMARK = true;
+    }
+    if(argc > 5 && strcmp(argv[5], "save_result") == 0) {
+        SAVE_RESULT = true;        
+    }
 
     tk::dnn::Yolo3Detection yolo;
     tk::dnn::CenternetDetection cnet;
@@ -76,8 +83,9 @@ int main(int argc, char *argv[]) {
 
     cv::Mat frame;
     cv::Mat dnn_input;
+    if(!BENCHMARK) {
     cv::namedWindow("detection", cv::WINDOW_NORMAL);
-    
+    }
     std::vector<tk::dnn::box> detected_bbox;
 
     while(gRun) {
@@ -91,9 +99,13 @@ int main(int argc, char *argv[]) {
         
         //inference
         detNN->update(dnn_input);
+        if(!BENCHMARK) {
         frame = detNN->draw(frame);
+        }
 
+        if(!BENCHMARK) {
         cv::imshow("detection", frame);
+        }
         cv::waitKey(1);
         if(SAVE_RESULT)
             resultVideo << frame;
@@ -106,9 +118,9 @@ int main(int argc, char *argv[]) {
     std::cout<<"Min: "<<*std::min_element(detNN->stats.begin(), detNN->stats.end())<<" ms\n";    
     std::cout<<"Max: "<<*std::max_element(detNN->stats.begin(), detNN->stats.end())<<" ms\n";    
     for(int i=0; i<detNN->stats.size(); i++) mean += detNN->stats[i]; mean /= detNN->stats.size();
-    std::cout<<"Avg: "<<mean<<" ms\n"<<COL_END;   
+    std::cout<<"Avg: "<<mean<<" ms\n";
+    std::cout<<"Avg FPS: " << 1000 / mean <<COL_END;   
     
 
     return 0;
 }
-