Assets/GadGame/Scenes/Main.unity

@@ -38,7 +38,7 @@ RenderSettings:
   m_ReflectionIntensity: 1
   m_CustomReflection: {fileID: 0}
   m_Sun: {fileID: 0}
-  m_IndirectSpecularColor: {r: 0.37311953, g: 0.38074014, b: 0.35872743, a: 1}
+  m_IndirectSpecularColor: {r: 0.37311918, g: 0.3807398, b: 0.35872716, a: 1}
   m_UseRadianceAmbientProbe: 0
 --- !u!157 &3
 LightmapSettings:
@@ -153,7 +153,7 @@ MonoBehaviour:
   m_Name: 
   m_EditorClassIdentifier: 
   _graphApi: {fileID: 11400000, guid: c2740d22f4ae04448a082f5f49761822, type: 2}
-  _promotionId: 63fb40435c803000
+  _promotionId: 63c67c5434403000
   _guestUpdatedSubscription: {fileID: 11400000, guid: f98ac02dda5623c4c82d342ee9602420, type: 2}
 --- !u!4 &398718624
 Transform:

Assets/GadGame/Scripts/Coffee/CoffeeController.cs

@@ -6,7 +6,6 @@ using TMPro;
 using UnityEngine;
 using UnityEngine.Serialization;
 using UnityEngine.UI;
-using GadGame.Network;
 
 namespace GadGame.Scripts.Coffee
 {
@@ -33,10 +32,8 @@ namespace GadGame.Scripts.Coffee
         private bool _isLoading;
         private float _timer;
 
-        private async void Awake()
+        private void Awake()
         {
-            await P4PGraphqlManager.Instance.JoinPromotion();
-            await P4PGraphqlManager.Instance.SubmitGameSession(0);
             _idleBg.alpha = 1;
             _loading.transform.DOLocalRotate(new Vector3(0, 0, 360), 10 / _loadingSpeed, RotateMode.FastBeyond360)
                 .SetLoops(-1)
@@ -133,7 +130,7 @@ namespace GadGame.Scripts.Coffee
         
         private void SetReadyCountDown(float progress){
             // _hintText.text = _texts[1];
-            _process.fillAmount = 1 - progress ;
+            _process.fillAmount = 1- progress ;
         }
         
     }

Assets/GadGame/Scripts/LoadImageEncoded.cs

@@ -25,6 +25,8 @@ namespace GadGame.Scripts
             //Decode the Base64 string to a byte array
             byte[] imageBytes = UdpSocket.Instance.DataReceived.GenerateImageSuccess ? File.ReadAllBytes(streamingData) : Convert.FromBase64String(streamingData);
 
+            // byte[] imageBytes = Convert.FromBase64String(encodeString);
+
             _texture.LoadImage(imageBytes); // Automatically resizes the texture dimensions
             // _texture.Apply();
             var sprite = Sprite.Create(_texture, new Rect(0, 0, _texture.width, _texture.height), new Vector2(0.5f, 0.5f), 100);

Assets/GadGame/Scripts/MainFlow.cs

@@ -27,7 +27,6 @@ namespace GadGame
             string _macAddress = GetMacAddressString();
             Debug.Log(_macAddress);
             await P4PGraphqlManager.Instance.LoginMachine(_macAddress);
-            await P4PGraphqlManager.Instance.CreateGuest();
         }
 
         private  async void Start()

Assets/GadGame/Scripts/Network/P4PGraphqlManager.cs

@@ -183,7 +183,7 @@ namespace GadGame.Network
             var socket = await _graphApi.Subscribe(query);
             if (socket.State == WebSocketState.Open)
             {
-                var link = $"https://play4promo.online/brands/{_promotionId}/scan-qr?token={_userAccessToken}&img=";
+                var link = $"https://play4promo.online/brands/{_promotionId}/scan-qr?token={_userAccessToken}";
                 Debug.Log(link);
                 return EncodeTextToQrCode(link);
             }

Assets/GadGame/Scripts/State/MainFlowState/CTAState.cs

@@ -23,13 +23,7 @@ namespace GadGame.State.MainFlowState
         public override void Update(float time)
         {
             Runner.EncodeImage.Raise(UdpSocket.Instance.DataReceived.StreamingData);
-            if(_scanSuccess) 
-            {
-                UdpSocket.Instance.SendDataToPython("End");
-                Runner.SetState<IdleState>();
-                return;
-            }
-
+            if(_scanSuccess) return;
             if (!UdpSocket.Instance.DataReceived.Engage)
             {
                 _leaveTimer += Time.deltaTime;

Assets/StreamingAssets/MergeFace/__pycache__/backup_FD.cpython-311.pyc.meta

@@ -1,7 +0,0 @@
-fileFormatVersion: 2
-guid: 5d25199b74a8a56878b29d2ae14fac9f
-DefaultImporter:
-  externalObjects: {}
-  userData: 
-  assetBundleName: 
-  assetBundleVariant: 

Assets/StreamingAssets/MergeFace/backup_FD.py

@@ -1,48 +0,0 @@
-from mtcnn import MTCNN
-from facenet_pytorch import MTCNN
-import torch
-import math
-import cv2
-import numpy as np
-
-K_MULTIPLIER = 1.2
-
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
-detector = MTCNN(keep_all=True, device=device)
-
-
-class FaceDetection:
-    def __init__(self,):
-        self.detector = detector
-
-    def calculate_distance(self, p1, p2) -> float:
-        return math.sqrt((p1[0] - p2[0]) ** 2 + (p1[1] - p2[1]) ** 2)
-
-    def calculate_dis_to_cp(self, cx, cy, face_cx, face_cy) -> float:
-        return math.sqrt((face_cx - cx) ** 2 + (face_cy - cy) ** 2)
-
-    def detect_face(self, frame, cx, cy) -> bool:
-        boxes, probs, landmarks = self.detector.detect(frame, landmarks=True)
-
-        for box, landmark in zip(boxes, landmarks):
-            # Draw bounding box
-            x1, y1, x2, y2 = map(int, box)
-
-            face_cx = int(x1 + (x2 - x1) / 2)
-            face_cy = int(y1 + (y2 - y1) / 2)
-
-            if len(landmark) >= 5:
-                nose = landmark[2]
-                left_eye = landmark[0]
-                right_eye = landmark[1]
-
-                # Calculate distances
-                distance_left = self.calculate_distance(nose, left_eye)
-                distance_right = self.calculate_distance(nose, right_eye)
-
-                # Check if distances exceed threshold
-                if not (distance_left > K_MULTIPLIER * distance_right or distance_right > K_MULTIPLIER * distance_left or
-                        self.calculate_dis_to_cp(cx, cy, face_cx, face_cy) > 30):
-                    return True
-
-        return False

Assets/StreamingAssets/MergeFace/backup_FD.py.meta

@@ -1,7 +0,0 @@
-fileFormatVersion: 2
-guid: f4f9204610747ca8bbc62f2c4c03f4b7
-DefaultImporter:
-  externalObjects: {}
-  userData: 
-  assetBundleName: 
-  assetBundleVariant: 

Assets/StreamingAssets/MergeFace/demo.py

@@ -51,22 +51,13 @@ class MainProgram:
 
         self.client = NovitaClient("48cc2b16-286f-49c8-9581-f409b68359c4")
 
-        self.client_minio = Minio("192.168.1.186:50047",
-                                  access_key="play4promo_user",
-                                  secret_key="12345678",
-                                  secure=False
-                                  )
-
-        self.bucket_name = "play4promo"
-        self.des_file = "sid/final_image.jpg"
-
         self.ready_success = False
 
         self.show_success = False
 
         self.check_save, self.check_generate = False, False
 
-        self.forward_face = FaceDetection()
+        self.forward_face = Face_detection.FaceDetection()
 
     def convertFrame(self, frame) -> str:
         encode_param = [int(cv2.IMWRITE_JPEG_QUALITY), 90]
@@ -86,18 +77,13 @@ class MainProgram:
 
     def generate_image(self):
         image_url, des = self.get_image()
-        image_path = "./image/merge_face.png"
 
         res = self.client.merge_face(
             image=image_url,
             face_image="./image/output.jpg",
         )
 
-        base64_to_image(res.image_file).save(image_path)
-
-        self.client_minio.fput_object(
-            self.bucket_name, self.des_file, image_path,
-        )
+        base64_to_image(res.image_file).save("./image/merge_face.png")
 
         self.send_data_unity["Description"] = des
         self.send_data_unity["GenerateImageSuccess"] = True
@@ -126,72 +112,143 @@ class MainProgram:
 
         return np.concatenate((bbox, conf[:, np.newaxis]), axis=1)
 
-    def check_engage(self, frame):
+    def get_person_bbox(self, frame):
         # Perform object detection with YOLOv8 class = 0 indicate person class
         outs = self.person_model(frame, classes=[0], conf=0.7)
         if not outs[0].boxes.xyxy.tolist():
-            return False
+            detection = np.empty((0, 6))
+        # Extract relevant information from detections for boxmot
         else:
+            boxes = outs[0].boxes.xyxy.tolist()
+            classes = outs[0].boxes.cls.tolist()
+            confidences = outs[0].boxes.conf.tolist()
+            detection = np.array([box + [conf, cls] for box, conf, cls in zip(boxes, confidences, classes)])
+
+        return detection
+
+    def check_engage(self, x1, x2) -> bool:
+        if not (x1 > self.red_zone_width[1] or x2 < self.red_zone_width[0]):
             return True
 
+        return False
+
     def cropped_image(self, frame, x1, y1, x2, y2):
         return frame[y1: y2, x1: x2]
 
     def calculate_dis_to_cp(self, cx, cy, face_cx, face_cy) -> float:
         return math.sqrt((face_cx - cx) ** 2 + (face_cy - cy) ** 2)
 
-    def check_ready(self, frame):
-        return self.forward_face.detect_face(frame, self.face_zone_center_point[0],
-                                             self.face_zone_center_point[1])
+    def check_ready(self, x1, y1, x2, y2, frame):
+        person_frame = self.cropped_image(frame, x1, y1, x2, y2)
+
+        # out = self.face_model(person_frame)
+        # results = sv.Detections.from_ultralytics(out[0])
+        # bbox = results.xyxy.astype(np.int_)
+        #
+        # face_cx, face_cy = (int(bbox[0][0] + x1 + (bbox[0][2] - bbox[0][0]) / 2),
+        #                     int(bbox[0][1] + y1 + (bbox[0][3] - bbox[0][1]) / 2))
+        #
+        # dis = self.calculate_dis_to_cp()
+
+        return self.forward_face.detect_face(person_frame, self.face_zone_center_point[0],
+                                             self.face_zone_center_point[1], x1, y1)
 
     def person_process(self, frame):
         # Perform person detection
-        self.send_data_unity["PassBy"] = self.check_engage(frame)
-        self.send_data_unity["Engage"] = self.send_data_unity["PassBy"]
+        person_detections = self.get_person_bbox(frame)
 
-        received_data = self.sock.ReadReceivedData()
+        # Update the tracker with person detections
+        tracked_objects = self.tracker.update(person_detections, frame)
 
-        if received_data == "Begin":
-            self.ready_success = True
+        track_list = []
 
-        elif received_data == "End":
-            self.ready_success = False
-            self.check_save = False
-            self.check_generate = False
-            os.remove("./image/output.jpg")
-            os.remove("./image/merge_face.png")
-            self.send_data_unity: dict = {
-                "PassBy": False,
-                "Engage": False,
-                "Ready": False,
-                "Gender": None,
-                "AgeMin": None,
-                "AgeMax": None,
-                "GenerateImageSuccess": False,
-                "Description": ""
-            }
+        frame_to_crop = frame.copy()
 
-        if not self.ready_success:
-            self.send_data_unity["Ready"] = True if self.check_ready(frame) else False
+        engage = False
 
-        elif not self.check_save:
-            cv2.imwrite("./image/output.jpg", self.cropped_image(frame, self.face_zone_width[0],
-                                                                 self.face_zone_height[0],
-                                                                 self.face_zone_width[1],
-                                                                 self.face_zone_height[1]))
-            self.check_save = True
+        for track in tracked_objects.astype(int):
+            x1, y1, x2, y2, track_id, conf, cls, _ = track
+            track_list.append(track_id)
+            # cv2.rectangle(self.frame_to_show, (x1, y1), (x2, y2), (0, 255, 0), 2)
+            # cv2.putText(self.frame_to_show, f"ID: {track_id} Conf: {conf:.2f}", (x1, y1 - 10),
+            #             cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
+            #
+            # cv2.rectangle(self.frame_to_show, (self.red_zone_width[0], self.red_zone_height[0]),
+            #               (self.red_zone_width[1], self.red_zone_height[1]), (255, 0, 0), 2)
 
-        elif not self.check_generate:
-            if str(self.send_data_unity["Gender"]) == "None":
-                self.predict_age_and_gender()
+            if not engage:
+                engage = self.check_engage(x1, x2)
+
+            if not self.focus_id:
+                self.focus_id = track_id if self.check_ready(x1, y1, x2, y2, frame_to_crop) else None
+
+            elif track_id != self.focus_id:
+                continue
 
             else:
-                self.generate_image()
-                self.check_generate = True
+                received_data = self.sock.ReadReceivedData()
 
-        elif self.show_success:
-            self.check_save = False
-            self.check_generate = False
+                if received_data == "Begin":
+                    self.ready_success = True
+
+                elif received_data == "End":
+                    self.ready_success = False
+                    self.check_save = False
+                    self.check_generate = False
+                    os.remove("./image/output.jpg")
+                    os.remove("./image/merge_face.png")
+                    self.send_data_unity: dict = {
+                        "PassBy": False,
+                        "Engage": False,
+                        "Ready": False,
+                        "Gender": None,
+                        "AgeMin": None,
+                        "AgeMax": None,
+                        "GenerateImageSuccess": False,
+                        "Description": ""
+                    }
+
+                if not self.ready_success:
+                    self.send_data_unity["Ready"] = True if self.check_ready(x1, y1, x2, y2, frame_to_crop) else False
+
+                elif not self.check_save:
+                    cv2.imwrite("./image/output.jpg", self.cropped_image(frame, x1, y1, x2, y2))
+                    self.check_save = True
+
+                elif not self.check_generate:
+                    if str(self.send_data_unity["Gender"]) == "None":
+                        self.predict_age_and_gender()
+
+                    else:
+                        self.generate_image()
+                        self.check_generate = True
+
+                elif self.show_success:
+                    self.check_save = False
+                    self.check_generate = False
+
+        if track_list:
+            self.send_data_unity["PassBy"] = True
+            self.send_data_unity["Engage"] = engage
+
+        else:
+            self.send_data_unity["Engage"] = False
+            self.send_data_unity["PassBy"] = False
+            self.send_data_unity["Ready"] = False
+
+        if self.focus_id not in track_list:
+            if self.frame_count_remove_idx == 20:
+                self.frame_count_remove_idx = 0
+                self.focus_id = None
+
+            else:
+                self.frame_count_remove_idx += 1
+
+        else:
+            self.frame_count_remove_idx = 0
+
+        # cv2.putText(self.frame_to_show, f"Focus id: {self.focus_id}", (20, 20), cv2.FONT_HERSHEY_SIMPLEX,
+        #             1.0, (0, 255, 255), 2)
 
     def __call__(self):
         cap = cv2.VideoCapture(0)
@@ -223,8 +280,6 @@ class MainProgram:
             frame_to_handle = frame.copy()
             self.frame_to_show = frame.copy()
 
-            # self.person_process(frame_to_handle)
-
             try:
                 self.person_process(frame_to_handle)
 

Assets/StreamingAssets/MergeFace/lib.py

@@ -32,8 +32,3 @@ import os
 from facenet_pytorch import MTCNN
 import torch
 import math
-
-from backup_FD import FaceDetection
-
-from minio import Minio
-from minio.error import S3Error