add face recognize app

examples/vision/ai_vision/nn_face_recognize.py

@@ -1,100 +1,4 @@
-from maix import nn, camera, display, image, time, touchscreen, app
-import os
-import math
-
-
-recognizer = nn.FaceRecognizer(detect_model="/root/models/retinaface.mud", feature_model = "/root/models/face_feature.mud")
-
-# if os.path.exists("/root/faces.bin"):
-#     recognizer.load_faces("/root/faces.bin")
-
-cam = camera.Camera(recognizer.input_width(), recognizer.input_height(), recognizer.input_format())
-disp = display.Display()
-ts = touchscreen.TouchScreen()
-back_btn_pos = (0, 0, 70, 30) # x, y, w, h
-learn_btn_pos = (0, recognizer.input_height() - 30, 60, 30)
-clear_btn_pos = (recognizer.input_width() - 60, recognizer.input_height() - 30, 60, 30)
-back_btn_disp_pos = image.resize_map_pos(cam.width(), cam.height(), disp.width(), disp.height(), image.Fit.FIT_CONTAIN, back_btn_pos[0], back_btn_pos[1], back_btn_pos[2], back_btn_pos[3])
-learn_btn_disp_pos = image.resize_map_pos(cam.width(), cam.height(), disp.width(), disp.height(), image.Fit.FIT_CONTAIN, learn_btn_pos[0], learn_btn_pos[1], learn_btn_pos[2], learn_btn_pos[3])
-clear_btn_disp_pos = image.resize_map_pos(cam.width(), cam.height(), disp.width(), disp.height(), image.Fit.FIT_CONTAIN, clear_btn_pos[0], clear_btn_pos[1], clear_btn_pos[2], clear_btn_pos[3])
-pressed_flag = [False, False, False]
-
-def draw_btns(img : image.Image):
-    img.draw_rect(back_btn_pos[0], back_btn_pos[1], back_btn_pos[2], back_btn_pos[3], image.Color.from_rgb(255, 255, 255), 2)
-    img.draw_string(back_btn_pos[0] + 4, back_btn_pos[1] + 8, "< back", image.COLOR_WHITE)
-    img.draw_rect(learn_btn_pos[0], learn_btn_pos[1], learn_btn_pos[2], learn_btn_pos[3], image.Color.from_rgb(255, 255, 255), 2)
-    img.draw_string(learn_btn_pos[0] + 4, learn_btn_pos[1] + 8, "learn", image.COLOR_WHITE)
-    img.draw_rect(clear_btn_pos[0], clear_btn_pos[1], clear_btn_pos[2], clear_btn_pos[3], image.Color.from_rgb(255, 255, 255), 2)
-    img.draw_string(clear_btn_pos[0] + 4, clear_btn_pos[1] + 8, "clear", image.COLOR_WHITE)
-
-def is_in_button(x, y, btn_pos):
-    return x > btn_pos[0] and x < btn_pos[0] + btn_pos[2] and y > btn_pos[1] and y < btn_pos[1] + btn_pos[3]
-
-def on_touch(x, y, pressed):
-    '''
-        Return learn, clear, ret
-    '''
-    global pressed_flag
-    if pressed:
-        if is_in_button(x, y, back_btn_disp_pos):
-            pressed_flag[2] = True
-        elif is_in_button(x, y, learn_btn_disp_pos):
-            pressed_flag[0] = True
-        elif is_in_button(x, y, clear_btn_disp_pos):
-            pressed_flag[1] = True
-        else: # cancel
-            pressed_flag = [False, False, False]
-    else:
-        if pressed_flag[0]:
-            print("learn btn click")
-            pressed_flag[0] = False
-            return True, False, False
-        if pressed_flag[1]:
-            print("clear btn click")
-            pressed_flag[1] = False
-            return False, True, False
-        if pressed_flag[2]:
-            print("back btn click")
-            pressed_flag[2] = False
-            return False, False, True
-    return False, False, False
-
-# Init key will cancel the default ok button function(exit app)
-learn_id = 0
-last_learn_img = None
-last_learn_t = 0
-
-while not app.need_exit():
-    x, y, pressed = ts.read()
-    learn, clear, back = on_touch(x, y, pressed)
-    if back:
-        break
-    elif clear:
-        for i in range(len(recognizer.labels) - 1):
-            recognizer.remove_face(0)
-    img = cam.read()
-    faces = recognizer.recognize(img, 0.5, 0.45, 0.8, learn, learn)
-    for obj in faces:
-        color = image.COLOR_RED if obj.class_id == 0 else image.COLOR_GREEN
-        img.draw_rect(obj.x, obj.y, obj.w, obj.h, color = color)
-        radius = math.ceil(obj.w / 10)
-        img.draw_keypoints(obj.points, color, size = radius if radius < 5 else 4)
-        msg = f'{recognizer.labels[obj.class_id]}: {obj.score:.2f}'
-        img.draw_string(obj.x, obj.y, msg, color = color)
-        if learn and obj.class_id == 0: # unknown face, we add it
-            name = f"id_{learn_id}"
-            print("add face:", name)
-            recognizer.add_face(obj, name)
-            learn_id += 1
-        if learn:
-            last_learn_img = obj.face
-            last_learn_t = time.ticks_s()
-    # show learned face on left-top
-    if last_learn_img and time.ticks_s() - last_learn_t < 5:
-        img.draw_image(0, 0, last_learn_img)
-    if learn:
-        recognizer.save_faces("/root/faces.bin")
-    draw_btns(img)
-    disp.show(img)
 
+# see https://github.com/sipeed/maixpy/blob/main/projects/app_face_recognizer
+# and https://wiki.sipeed.com/maixpy/doc/zh/vision/face_recognition.html
 

projects/app_face_recognizer/.gitignore

@@ -0,0 +1,5 @@
+
+build
+dist
+/CMakeLists.txt
+

projects/app_face_recognizer/app.yaml

@@ -0,0 +1,10 @@
+id: face_recognizer
+name: Face Recognizer
+version: 1.0.0
+author: Sipeed Ltd
+icon: face.json
+desc: Learn and recognize face.
+files:
+  - face.json
+  - app.yaml
+  - main.py

projects/app_face_recognizer/main.py

@@ -0,0 +1,111 @@
+from maix import nn, camera, display, image, time, touchscreen, app
+import math
+
+pressed_flag = [False, False, False]
+learn_id = 0
+
+def main(disp):
+    recognizer = nn.FaceRecognizer(detect_model="/root/models/retinaface.mud", feature_model = "/root/models/face_feature.mud")
+
+    # if os.path.exists("/root/faces.bin"):
+    #     recognizer.load_faces("/root/faces.bin")
+
+    cam = camera.Camera(recognizer.input_width(), recognizer.input_height(), recognizer.input_format())
+    ts = touchscreen.TouchScreen()
+    back_btn_pos = (0, 0, 70, 30) # x, y, w, h
+    learn_btn_pos = (0, recognizer.input_height() - 30, 60, 30)
+    clear_btn_pos = (recognizer.input_width() - 60, recognizer.input_height() - 30, 60, 30)
+    back_btn_disp_pos = image.resize_map_pos(cam.width(), cam.height(), disp.width(), disp.height(), image.Fit.FIT_CONTAIN, back_btn_pos[0], back_btn_pos[1], back_btn_pos[2], back_btn_pos[3])
+    learn_btn_disp_pos = image.resize_map_pos(cam.width(), cam.height(), disp.width(), disp.height(), image.Fit.FIT_CONTAIN, learn_btn_pos[0], learn_btn_pos[1], learn_btn_pos[2], learn_btn_pos[3])
+    clear_btn_disp_pos = image.resize_map_pos(cam.width(), cam.height(), disp.width(), disp.height(), image.Fit.FIT_CONTAIN, clear_btn_pos[0], clear_btn_pos[1], clear_btn_pos[2], clear_btn_pos[3])
+
+    def draw_btns(img : image.Image):
+        img.draw_rect(back_btn_pos[0], back_btn_pos[1], back_btn_pos[2], back_btn_pos[3], image.Color.from_rgb(255, 255, 255), 2)
+        img.draw_string(back_btn_pos[0] + 4, back_btn_pos[1] + 8, "< back", image.COLOR_WHITE)
+        img.draw_rect(learn_btn_pos[0], learn_btn_pos[1], learn_btn_pos[2], learn_btn_pos[3], image.Color.from_rgb(255, 255, 255), 2)
+        img.draw_string(learn_btn_pos[0] + 4, learn_btn_pos[1] + 8, "learn", image.COLOR_WHITE)
+        img.draw_rect(clear_btn_pos[0], clear_btn_pos[1], clear_btn_pos[2], clear_btn_pos[3], image.Color.from_rgb(255, 255, 255), 2)
+        img.draw_string(clear_btn_pos[0] + 4, clear_btn_pos[1] + 8, "clear", image.COLOR_WHITE)
+
+    def is_in_button(x, y, btn_pos):
+        return x > btn_pos[0] and x < btn_pos[0] + btn_pos[2] and y > btn_pos[1] and y < btn_pos[1] + btn_pos[3]
+
+    def on_touch(x, y, pressed):
+        '''
+            Return learn, clear, ret
+        '''
+        global pressed_flag, learn_id
+        if pressed:
+            if is_in_button(x, y, back_btn_disp_pos):
+                pressed_flag[2] = True
+            elif is_in_button(x, y, learn_btn_disp_pos):
+                pressed_flag[0] = True
+            elif is_in_button(x, y, clear_btn_disp_pos):
+                pressed_flag[1] = True
+            else: # cancel
+                pressed_flag = [False, False, False]
+        else:
+            if pressed_flag[0]:
+                print("learn btn click")
+                pressed_flag[0] = False
+                return True, False, False
+            if pressed_flag[1]:
+                print("clear btn click")
+                pressed_flag[1] = False
+                learn_id = 0
+                return False, True, False
+            if pressed_flag[2]:
+                print("back btn click")
+                pressed_flag[2] = False
+                return False, False, True
+        return False, False, False
+
+    # Init key will cancel the default ok button function(exit app)
+    last_learn_img = None
+    last_learn_t = 0
+
+    while not app.need_exit():
+        x, y, pressed = ts.read()
+        learn, clear, back = on_touch(x, y, pressed)
+        if back:
+            break
+        elif clear:
+            for i in range(len(recognizer.labels) - 1):
+                recognizer.remove_face(0)
+        img = cam.read()
+        faces = recognizer.recognize(img, 0.5, 0.45, 0.8, learn, learn)
+        for obj in faces:
+            color = image.COLOR_RED if obj.class_id == 0 else image.COLOR_GREEN
+            img.draw_rect(obj.x, obj.y, obj.w, obj.h, color = color)
+            radius = math.ceil(obj.w / 10)
+            img.draw_keypoints(obj.points, color, size = radius if radius < 5 else 4)
+            msg = f'{recognizer.labels[obj.class_id]}: {obj.score:.2f}'
+            img.draw_string(obj.x, obj.y - 10, msg, color = color)
+            if learn and obj.class_id == 0: # unknown face, we add it
+                name = f"id_{learn_id}"
+                print("add face:", name)
+                recognizer.add_face(obj, name)
+                learn_id += 1
+            if learn:
+                last_learn_img = obj.face
+                last_learn_t = time.ticks_s()
+        # show learned face on left-top
+        if last_learn_img and time.ticks_s() - last_learn_t < 5:
+            img.draw_image(0, 0, last_learn_img)
+        if learn:
+            recognizer.save_faces("/root/faces.bin")
+        draw_btns(img)
+        disp.show(img)
+
+
+disp = display.Display()
+try:
+    main(disp)
+except Exception:
+    import traceback
+    msg = traceback.format_exc()
+    img = image.Image(disp.width(), disp.height())
+    img.draw_string(0, 0, msg, image.COLOR_WHITE)
+    disp.show(img)
+    while not app.need_exit():
+        time.sleep_ms(100)

projects/app_yolov8_seg/app.yaml

@@ -3,7 +3,7 @@ name: YOLOv8 Segmentation
 version: 1.0.0
 author: Sipeed Ltd
 icon: icon.png
-desc: Detect human body 17 keypoints.
+desc: YOLOv8 detect objects and segment object
 files:
   - icon.png
   - app.yaml