Merge branch 'main' of https://github.com/aaj2005/MAAL-AstonHack9

AI/SpeedTesting.py

@@ -1,27 +1,28 @@
 import cv2
 import numpy as np
-from keras.models import model_from_json
 import time
+from keras.models import model_from_json
 
 emotion_dict = {0: "Angry", 1: "Disgusted", 2: "Fearful", 3: "Happy", 4: "Neutral", 5: "Sad", 6: "Surprised"}
 
+cumulativeEmotionArray = [0] *7
 
 # load json and create model
-json_file = open("AI\model\emotion_model.json", 'r')
+json_file = open("AI/model/emotion_model.json", 'r')
 loaded_model_json = json_file.read()
 json_file.close()
 emotion_model = model_from_json(loaded_model_json)
 
 # load weights into new model
-emotion_model.load_weights("AI\model\emotion_model.h5")
+emotion_model.load_weights("AI/model/emotion_model.h5")
 print("Loaded model from disk")
 
 def colourBasedEmotion(emotion):
     colourDict = {0: (0, 0, 255), 1: (0, 255, 0), 2: (255, 0, 255), 3: (0, 255, 255), 4: (255,255,255), 5: (255, 0, 0), 6: (0, 165, 255)}
     return colourDict[emotion]
 
 def takeSinglePhoto(frame):
-    face_detector = cv2.CascadeClassifier('AI\haarcascade_frontalface_default.xml')
+    face_detector = cv2.CascadeClassifier('AI/haarcascade_frontalface_default.xml')
     gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
     num_faces = face_detector.detectMultiScale(gray_frame, scaleFactor=1.3, minNeighbors=5)
     maxindex = -1
@@ -36,33 +37,39 @@ def takeSinglePhoto(frame):
         # predict the emotions
         emotion_prediction = emotion_model.predict(cropped_img)
         maxindex = int(np.argmax(emotion_prediction))
+        if (maxindex != -1):
+            cumulativeEmotionArray[maxindex] += 1
     return [maxindex, x, y, w, h]
 
+
+def getCumulativeArray():
+    return cumulativeEmotionArray
+
 def mainLoop(frame):
 
-    arrayOfEmotions = [0,0,0,0,0,0,0]
+    # arrayOfEmotions = [0,0,0,0,0,0,0]
     frame = cv2.resize(frame, (1024, 576))
-    while True:
-        # Find haar cascade to draw bounding box around face
-        maxindex, x, y, w, h = takeSinglePhoto(frame)
-        if x != -1 and y != -1 and w != -1 and h != -1:
-            frame = cv2.rectangle(frame, (x, y-50), (x+w, y+h+10), colourBasedEmotion(maxindex), 4)
-            if maxindex != -1:
-                frame = cv2.putText(frame, emotion_dict[maxindex], (x+5, y-20), cv2.FONT_HERSHEY_SIMPLEX, 1, colourBasedEmotion(maxindex), 2, cv2.LINE_AA)
-            arrayOfEmotions[maxindex] +=1     
-        # If more than 5 seconds have elapsed
-        # if needArray:
-        #     return arrayOfEmotions
-        cv2.imshow('Emotion Detection', frame)
-        if cv2.waitKey(1) & 0xFF == ord('q'):
-            break
-        img_encode = cv2.imencode('.jpg', frame)[1] 
-        data_encode = np.array(img_encode) 
-        # byte_encode = data_encode.tobytes() 
-        return img_encode
+    # while True:
+    #     # Find haar cascade to draw bounding box around face
+    return takeSinglePhoto(frame)
+    #     if x != -1 and y != -1 and w != -1 and h != -1:
+    #         frame = cv2.rectangle(frame, (x, y-50), (x+w, y+h+10), colourBasedEmotion(maxindex), 4)
+    #         if maxindex != -1:
+    #             frame = cv2.putText(frame, emotion_dict[maxindex], (x+5, y-20), cv2.FONT_HERSHEY_SIMPLEX, 1, colourBasedEmotion(maxindex), 2, cv2.LINE_AA)
+    #         arrayOfEmotions[maxindex] +=1     
+    #     # If more than 5 seconds have elapsed
+    #     # if needArray:
+    #     #     return arrayOfEmotions
+    #     cv2.imshow('Emotion Detection', frame)
+    #     if cv2.waitKey(1) & 0xFF == ord('q'):
+    #         break
+    #     img_encode = cv2.imencode('.jpg', frame)[1] 
+    #     data_encode = np.array(img_encode) 
+    #     # byte_encode = data_encode.tobytes() 
+    #     return img_encode
 
 
-cap = cv2.VideoCapture(0)
-while True:
-    ret, frame = cap.read()
-    print(mainLoop(frame))
+# cap = cv2.VideoCapture(0)
+# while True:
+#     ret, frame = cap.read()
+    # print(mainLoop(frame))

Pipfile

@@ -12,6 +12,8 @@ pillow = "*"
 numpy = "*"
 opencv-python-headless = "*"
 flask-cors = "*"
+keras = "*"
+tensorflow = "*"
 
 [dev-packages]
 

app.py

@@ -1,12 +1,16 @@
+from AI.SpeedTesting import mainLoop, getCumulativeArray
+from APIStuff.OpenAPI.initialPrompt import main
 from flask import Flask, request
-from flask_cors import CORS, cross_origin
+from flask_cors import CORS
+from io import BytesIO
+import json
 
-import cv2
 import base64
 import numpy as np
-from scipy import misc
 
-#something
+
+
+#somethin
 app = Flask(__name__)
 
 CORS(app)
@@ -25,41 +29,88 @@ def webp_to_jpg(webp_base64_string, output_file_path='output.jpg'):
     # Open WebP image using Pillow
     with Image.open(BytesIO(binary_data)) as webp_image:
         # Save as JPG
-        webp_image.convert('RGB').save(output_file_path, format='JPEG')
+        webp_image.convert('RGB').save(output_file_path, format='JPEG') 
 
 # Example usage
 
-
 @app.route("/", methods=["POST","GET"])
 def index():
     if request.method == "GET":
-        return "hello world indeed"
-	# print(test)
-	# print(request.headers)
+        return "hello world"
+
     data = request.json["imageData"]
-    webp_to_jpg(data)
-	# print(jpg_original)
-	# img = cv2.imdecode(jpg_as_np, flags=1)
-	# print(img)
-	# cv2.imwrite('./images.webp', img)
+    image = Image.open(BytesIO(data))
+    numpy_array = np.array(image)
+
+    webp_to_jpg(numpy_array)
+
 
     return {"1":"hello world"}
 
 
-@app.route("/api/frame", methods=["POST","GET"])
+
+
+@app.route("/api/frame", methods=["POST"])
 def imageProcess():
-    if request.method == "GET":
-        return "hello world"
-	# print(test)
-	# print(request.headers)
-    data = request.json["imageData"]
-    webp_to_jpg(data)
-	# print(jpg_original)
-	# img = cv2.imdecode(jpg_as_np, flags=1)
-	# print(img)
-	# cv2.imwrite('./images.webp', img)
+
+    base64_string = request.json["imageData"]
+    _, data = base64_string.split(',', 1)
+    # Decode the base64 data
+    decoded_data = base64.b64decode(data)
+    # Create a PIL Image
+    image = Image.open(BytesIO(decoded_data))
 
-    return {"1":"hello world"}
+    with BytesIO() as jpeg_buffer:
+        image.save(jpeg_buffer, format="JPEG")
+        jpeg_buffer.seek(0)
+        jpeg_image = Image.open(jpeg_buffer)
+
+        # Now convert the JPEG image to a NumPy array
+        numpy_array = np.array(jpeg_image)
+
+        # Here, numpy_array is ready and you can use it with mainLoop or any other function
+    newImg = list(map(str, mainLoop(numpy_array)))
+    img_str = json.dumps(newImg)
+
+    # Send as JSON
+    return img_str
+
+
+@app.route("/api/emotionsArray", methods=["POST"])
+def getEmotionsArray():
+
+    array = getCumulativeArray()
+    array = main()
+
+    newImg = list(map(str, array))
+    img_str = json.dumps(newImg)
+
+    return img_str
+
+@app.route("/api/promt", methods=["POST"])
+def getGPTOutput():
+
+    # Get array from json request - might not work ahahahh loser
+    jsonResponse = request.json()
+
+    # Parse the JSON data
+    parsedData = json.loads(jsonResponse)
+
+    # Access the array
+    dataArray = parsedData['data']
+
+    prompt = main(dataArray)
+
+    # newImg = list(map(str, array))
+    # img_str = json.dumps(newImg)
+
+
+
+    return img_str
+
+
+
+
 
 if __name__ == "__main__":
     app.run(port=2223)

frontend/package.json

@@ -12,6 +12,7 @@
     "react-router-dom": "^6.22.1",
     "react-scripts": "5.0.1",
     "react-webcam": "^7.2.0",
+    "socket.io-client": "^4.7.4",
     "web-vitals": "^2.1.4"
   },
   "scripts": {

frontend/src/components/BoundingBox.js

@@ -0,0 +1,38 @@
+import { useState, useEffect } from 'react';
+
+const BoundingBox = ({ boxData }) => {
+    const [ignore, x, y, width, height] = boxData;
+
+    const emotion_dict = {0: "Angry", 1: "Disgusted", 2: "Fearful", 3: "Happy", 4: "Neutral", 5: "Sad", 6: "Surprised"}
+    const colour_dict = {0: "red", 1: "green", 2: "pink", 3: "yellow", 4: "white", 5: "blue", 6: "orange"}
+    const colour = colour_dict[ignore]  
+    const textStyle = {
+      position: 'absolute',
+      top: `${y-25}px`,
+      left: `${x}px`,
+      display: 'flex',
+      justifyContent: 'center',
+      alignItems: 'center',
+      color: colour,  // Customize text color
+      fontSize: '16px',  // Customize font size
+      fontWeight: 'bold',  // Customize font weight
+    };
+
+
+    return (
+      <div>
+        <div className='bounding-box'
+          style={{
+            top: `${y}px`,
+            left: `${x}px`,
+            left: `calc(8vw + ${x}px)`,
+            width: `${width}px`,
+            height: `${height}px`,
+            borderColor: colour,
+          }}
+        ></div>
+        <div className='bounding-text' style={textStyle}>{emotion_dict[ignore]}</div>
+      </div>
+    );
+  };
+export default BoundingBox