vagheshpatel commited on
Commit
2f25f77
·
verified ·
1 Parent(s): 35d376a

Sync crowd-detection from metro-analytics-catalog

Browse files
Files changed (2) hide show
  1. README.md +10 -7
  2. export_and_quantize.sh +1 -1
README.md CHANGED
@@ -6,7 +6,7 @@
6
  |---|---|
7
  | **Category** | Object Detection (Crowd / Person Counting) |
8
  | **Source Framework** | PyTorch (Ultralytics) |
9
- | **Supported Precisions** | FP32, FP16, FP16-INT8 |
10
  | **Inference Engine** | OpenVINO |
11
  | **Hardware** | CPU, GPU, NPU |
12
  | **Detected Class** | `person` (COCO class 0) |
@@ -42,6 +42,10 @@ python3 -m venv .venv --system-site-packages
42
  source .venv/bin/activate
43
  ```
44
 
 
 
 
 
45
  ---
46
 
47
  ## Getting Started
@@ -148,17 +152,16 @@ The `export_and_quantize.sh` script downloads `test.jpg` automatically.
148
  Re-run the OpenVINO sample above.
149
  The script reads `test.jpg`, prints the crowd count to the console, and writes the annotated frame to `output.jpg`.
150
 
151
- Expected console output (representative -- actual count depends on the sample image):
152
 
153
  ```text
154
- Detected persons: <N>
155
  ```
156
 
157
- `output.jpg` is the same image with a green bounding box drawn around each detected person and the text `Crowd count: <N>` overlaid in the top-left corner.
158
 
159
- > **Tip:** The bundled `test.jpg` (an airport scene) is suitable for a quick
160
- > demo. For production testing, replace it with an image from your target
161
- > deployment site showing a representative crowd density.
162
 
163
  ### DLStreamer Sample
164
 
 
6
  |---|---|
7
  | **Category** | Object Detection (Crowd / Person Counting) |
8
  | **Source Framework** | PyTorch (Ultralytics) |
9
+ | **Supported Precisions** | FP32, FP16, INT8 (mixed-precision) |
10
  | **Inference Engine** | OpenVINO |
11
  | **Hardware** | CPU, GPU, NPU |
12
  | **Detected Class** | `person` (COCO class 0) |
 
42
  source .venv/bin/activate
43
  ```
44
 
45
+ > **Note:** The `--system-site-packages` flag is required so the virtual
46
+ > environment can access the system-installed OpenVINO and DLStreamer Python
47
+ > packages.
48
+
49
  ---
50
 
51
  ## Getting Started
 
152
  Re-run the OpenVINO sample above.
153
  The script reads `test.jpg`, prints the crowd count to the console, and writes the annotated frame to `output.jpg`.
154
 
155
+ Expected console output:
156
 
157
  ```text
158
+ Detected persons: 4
159
  ```
160
 
161
+ `output.jpg` is the same image with a green bounding box drawn around each detected person and the text `Crowd count: 4` overlaid in the top-left corner.
162
 
163
+ > **Tip:** For production testing, replace the bundled `test.jpg` with an image
164
+ > from your target deployment site showing a representative crowd density.
 
165
 
166
  ### DLStreamer Sample
167
 
export_and_quantize.sh CHANGED
@@ -38,7 +38,7 @@ fi
38
 
39
  echo "--- Downloading sample test image ---"
40
  if [[ ! -f test.jpg ]]; then
41
- wget -q -O test.jpg https://raw.githubusercontent.com/ultralytics/assets/main/im/airport.jpg
42
  echo "Downloaded: test.jpg"
43
  else
44
  echo "Already present: test.jpg"
 
38
 
39
  echo "--- Downloading sample test image ---"
40
  if [[ ! -f test.jpg ]]; then
41
+ wget -q -O test.jpg https://ultralytics.com/images/bus.jpg
42
  echo "Downloaded: test.jpg"
43
  else
44
  echo "Already present: test.jpg"