Update app.py

app.py

@@ -1,55 +1,72 @@
 import gradio as gr
 import cv2
-from huggingface_hub import hf_hub_download
 from gradio_webrtc import WebRTC
-from twilio.rest import Client
-import os
-from inference import YOLOv10
+import mediapipe as mp
+import time
 
-model_file = hf_hub_download(
-    repo_id="onnx-community/yolov10n", filename="onnx/model.onnx"
-)
+# 初始化 MediaPipe Hands
+mp_hands = mp.solutions.hands
+mp_drawing = mp.solutions.drawing_utils
+hands = mp_hands.Hands(min_detection_confidence=0.3, min_tracking_confidence=0.3)  # 降低置信度提升速度
 
-model = YOLOv10(model_file)
+# WebRTC 配置
+rtc_configuration = {
+    "iceServers": [{"urls": "stun:stun.l.google.com:19302"}],
+    "iceTransportPolicy": "relay"
+}
 
-account_sid = os.environ.get("TWILIO_ACCOUNT_SID")
-auth_token = os.environ.get("TWILIO_AUTH_TOKEN")
+# 控制每秒帧处理频率的时间
+last_process_time = time.time()
 
-if account_sid and auth_token:
-    client = Client(account_sid, auth_token)
+# 手势检测函数
+def detection(image, conf_threshold=0.5):
+    """
+    使用 MediaPipe Hands 进行手势检测。
+    """
+    global last_process_time
+    current_time = time.time()
+    
+    # 只每隔一定时间（比如0.1秒）才进行一次处理，减少计算负担
+    if current_time - last_process_time < 0.1:
+        return image  # 如果时间间隔太短，则直接返回原图像
 
-    token = client.tokens.create()
+    last_process_time = current_time
 
-    rtc_configuration = {
-        "iceServers": token.ice_servers,
-        "iceTransportPolicy": "relay",
-    }
-else:
-    rtc_configuration = None
+    # 将图像从 BGR 转换为 RGB（MediaPipe 需要 RGB 格式）
+    image_rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
 
+    # 将图像大小缩小到一个较小的尺寸，降低计算负担
+    image = cv2.resize(image, (640, 480))
 
-def detection(image, conf_threshold=0.3):
-    image = cv2.resize(image, (model.input_width, model.input_height))
-    new_image = model.detect_objects(image, conf_threshold)
-    return cv2.resize(new_image, (500, 500))
+    # 使用 MediaPipe Hands 处理图像
+    results = hands.process(image_rgb)
 
+    # 如果检测到手，绘制手部关键点
+    if results.multi_hand_landmarks:
+        for hand_landmarks in results.multi_hand_landmarks:
+            mp_drawing.draw_landmarks(
+                image, hand_landmarks, mp_hands.HAND_CONNECTIONS
+            )
 
-css = """.my-group {max-width: 600px !important; max-height: 600 !important;}
-                      .my-column {display: flex !important; justify-content: center !important; align-items: center !important};"""
+    # 返回带注释的图像
+    return image
 
+# Gradio 界面
+css = """.my-group {max-width: 600px !important; max-height: 600 !important;}
+         .my-column {display: flex !important; justify-content: center !important; align-items: center !important;}"""
 
 with gr.Blocks(css=css) as demo:
     gr.HTML(
         """
-    <h1 style='text-align: center'>
-    YOLOv10 Webcam Stream (Powered by WebRTC ⚡️)
-    </h1>
-    """
+        <h1 style='text-align: center'>
+        Hand Gesture Detection with MediaPipe (Powered by WebRTC ⚡️)
+        </h1>
+        """
     )
     gr.HTML(
         """
         <h3 style='text-align: center'>
-        <a href='https://arxiv.org/abs/2405.14458' target='_blank'>arXiv</a> | <a href='https://github.com/THU-MIG/yolov10' target='_blank'>github</a>
+        <a href='https://mediapipe.dev/'>MediaPipe Hands</a>
         </h3>
         """
     )
@@ -61,12 +78,11 @@ with gr.Blocks(css=css) as demo:
                 minimum=0.0,
                 maximum=1.0,
                 step=0.05,
-                value=0.30,
+                value=0.5,
             )
 
-        image.stream(
-            fn=detection, inputs=[image, conf_threshold], outputs=[image], time_limit=10
-        )
+        # 使用简化的stream函数，不使用queue参数
+        image.stream(fn=detection, inputs=[image, conf_threshold], outputs=[image], time_limit=10)
 
 if __name__ == "__main__":
     demo.launch()