Differences

This shows you the differences between two versions of the page.

--- products:sbc:edge2:npu:demos:retinaface [2023/09/19 22:42]
louis
+++ products:sbc:edge2:npu:demos:retinaface [2025/04/09 23:33] (current)
louis
@@ Line 1: / Line 1: @@
 ~~tag> NPU RetinaFace Edge2 PyTorch~~
-====== RetinaFace PyTorch VIM4 Demo - 5 ======
+====== RetinaFace PyTorch Edge2 Demo - 5 ======
 {{indexmenu_n>5}}
-===== Get Source Code =====
+===== Introduction =====
+RetinaFace is a face detection model. It can draw five key points on each face, including two eyes, nose and two corners of mouth.
+Inference results on Edge2.
+{{:products:sbc:edge2:npu:demos:retinaface-result.jpg?800|}}
+**Inference speed test**: USB camera about **39ms** per frame. MIPI camera about **33ms** per frame.
+===== Train Model =====
 The codes we use [[gh>bubbliiiing/retinaface-pytorch]].
@@ Line 136: / Line 146: @@
 ==== Get source code ====
-Clone the source code form our [[gh>khadas/edge2-npu]].
+Clone the source code from our [[gh>khadas/edge2-npu]].
 ```shell
@@ Line 172: / Line 182: @@
 $ bash build.sh
-# Run
+# Run USB camera
+$ cd install/retinaface_cap
+$ ./retinaface_cap data/model/retinaface.rknn usb 60
+# Run MIPI camera
 $ cd install/retinaface_cap
-$ ./retinaface_cap data/model/retinaface_cap.rknn 33
+$ ./retinaface_cap data/model/retinaface.rknn mipi 42
 ```
-''33'' is camera device index.
+<WRAP info >
+''60'' and ''42'' are camera device index.
+</WRAP>

Khadas Docs