diff --git a/integration/Window.py b/integration/Window.py
index 131033d3caf0c7b9c3d8556b81645fa4fda9768e..97fb6ba1bbc1cfa45202a4ea20a9d511dcbba7ef 100644
--- a/integration/Window.py
+++ b/integration/Window.py
@@ -23,6 +23,8 @@ from torch.utils.data import DataLoader, TensorDataset, random_split
import pickle
import joblib
+#this is the file for seperated hardware detectors
+
class Window:
def __init__(self, root):
self.root = root
diff --git a/integration/Window2-copyformlp.py b/integration/Window2-copyformlp.py
new file mode 100644
index 0000000000000000000000000000000000000000..e8c3beb2fbdc71d7baafd67f532aa49b5146bfe6
--- /dev/null
+++ b/integration/Window2-copyformlp.py
@@ -0,0 +1,1091 @@
+import threading
+import tkinter as tk
+from matplotlib.figure import Figure
+from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg
+from mpl_toolkits.mplot3d.art3d import Poly3DCollection
+import numpy as np
+import serial
+import serial.tools.list_ports
+from time import sleep
+import math
+import time
+import socket
+import os
+from PIL import Image, ImageTk
+from time import sleep, time
+import torch
+import torch.nn as nn
+import torch.optim as optim
+from sklearn.model_selection import train_test_split
+
+#this is the file for the combined hardware detector
+
+class Window:
+ def __init__(self, root):
+ self.input_port='COM9'
+ self.root = root
+ self.root.title("Integration")
+ self.ports = [port.device for port in serial.tools.list_ports.comports()]
+ self.arduino = None
+
+ # Set the initial size and position of the popup window
+ self.width = 1000
+ self.height = 600
+ screen_width = self.root.winfo_screenwidth()
+ screen_height = self.root.winfo_screenheight()
+ x = (screen_width // 2) - (self.width // 2)
+ y = (screen_height // 2) - (self.height // 2)
+ self.root.geometry(f"{self.width}x{self.height}+{x}+{y}")
+
+ # Configure the grid to be expandable
+ self.root.columnconfigure(0, weight=1)
+ self.root.columnconfigure(1, weight=1)
+ self.root.rowconfigure(0, weight=1)
+ self.root.rowconfigure(1, weight=1)
+
+ # Create a frame
+ self.frame1 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width / 3, height=self.height / 2)
+ self.frame1.grid(row=0, column=0, padx=10, pady=10, sticky="nsew")
+
+ self.frame2 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width * 2 / 3, height=self.height / 2)
+ self.frame2.grid(row=0, column=1, padx=10, pady=10, sticky="nsew")
+
+
+ self.frame3 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width / 3, height=self.height / 2)
+ self.frame3.grid(row=1, column=0, padx=10, pady=10, sticky="nsew")
+
+ self.frame4 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width * 2 / 3, height=self.height / 2)
+ self.frame4.grid(row=1, column=1, padx=10, pady=10, sticky="nsew")
+ self.frame4.grid_propagate(False)
+ label4 = tk.Label(self.frame4, text="Section 4")
+ label4.place(relx=0.5, rely=0.5, anchor='center')
+ self.start_button = tk.Button(self.frame2, text="Game Start", command=self.game_Start, width=15, height=1,
+ font=("Helvetica", 12))
+ self.start_button.place(relx=0.7, rely=0.15, anchor='center')
+
+ #self.imu_thread = threading.Thread(target=self.initial_IMU)
+ #self.emg_thread = threading.Thread(target=self._initialise_EMG_graph)
+ #self.emg_thread.start()
+ #self.imu_thread.start()
+
+
+
+
+ self.emg_data_1 = [-1.0] * 41
+ self.emg_data_2 = [-1.0] * 41
+
+ self.initial_IMU()
+ self._initialise_EMG_graph()
+ self.display_IMU_thread=threading.Thread(target=self.update_display)
+ self.display_EMG_thread=threading.Thread(target=self.EMG_Display)
+
+ # initial button, commands and 3D metrics for IMU detector
+ def initial_IMU(self):
+ # Serial Port Setup
+ if self.input_port in self.ports:#port maybe different on different laptop
+
+ self.label2 = tk.Label(self.frame2, text=f"Port: {self.input_port} ")
+ self.label2.place(relx=0.35, rely=0.8, anchor='center')
+ self.label1 = tk.Label(self.frame2,
+ text="click the Connect button to see the animation",
+ wraplength=self.width / 2)
+ self.label1.place(relx=0.5, rely=0.9, anchor='center')
+ # Add a button to start data transmission
+ self.start_buttonConnect = tk.Button(self.frame2, text="connect", command=self.start_data_transmission)
+ self.start_buttonConnect.place(relx=0.5, rely=0.8, anchor='center')
+
+ self.start_buttonDisConnect = tk.Button(self.frame2, text="Disconnect", command=self.disconnect)
+ self.start_buttonDisConnect.place(relx=0.7, rely=0.8, anchor='center')
+
+ else:
+ print("IMU is not connected")
+ self.label2 = tk.Label(self.frame2, text="Port: None ")
+ self.label2.place(relx=0.35, rely=0.8, anchor='center')
+ self.label1 = tk.Label(self.frame2,
+ text="Please check the IUM connection",
+ wraplength=self.width / 2)
+ self.label1.place(relx=0.5, rely=0.9, anchor='center')
+
+ sleep(1)
+
+ # Conversions
+ self.transmitting = False
+ self.toRad = 2 * np.pi / 360
+ self.toDeg = 1 / self.toRad
+
+ # Initialize Parameters
+ self.count = 0
+ self.averageroll = 0
+ self.averageyaw = 0
+ self.averagepitch = 0
+ self.averageemg = 0
+ self.iterations = 10 # EMG measurements to get average
+
+ # Create a figure for the 3D plot
+ self.fig = Figure(figsize=((self.width / 300), (self.height / 200)))
+ self.ax = self.fig.add_subplot(111, projection='3d')
+
+ # Set Limits
+ self.ax.set_xlim(-2, 2)
+ self.ax.set_ylim(-2, 2)
+ self.ax.set_zlim(-2, 2)
+
+ # Set labels
+ self.ax.set_xlabel('X')
+ self.ax.set_ylabel('Y')
+ self.ax.set_zlabel('Z',labelpad=0)
+
+ # Draw Axes
+ self.ax.quiver(0, 0, 0, 2, 0, 0, color='red', label='X-Axis', arrow_length_ratio=0.1) # X Axis (Red)
+ self.ax.quiver(0, 0, 0, 0, -2, 0, color='green', label='Y-Axis', arrow_length_ratio=0.1) # Y Axis (Green)
+ self.ax.quiver(0, 0, 0, 0, 0, 4, color='blue', label='Z-Axis', arrow_length_ratio=0.1) # Z Axis (Blue)
+
+ # Draw the board as a rectangular prism (solid)
+ self.prism_vertices = np.array([
+ [-1.5, -1, 0], [1.5, -1, 0], [1.5, 1, 0], [-1.5, 1, 0], # bottom vertices
+ [-1.5, -1, 0.1], [1.5, -1, 0.1], [1.5, 1, 0.1], [-1.5, 1, 0.1]
+ # top vertices (height=0.1 for visual thickness)
+ ])
+
+ self.prism_faces = [
+ [self.prism_vertices[j] for j in [0, 1, 2, 3]], # bottom face
+ [self.prism_vertices[j] for j in [4, 5, 6, 7]], # top face
+ [self.prism_vertices[j] for j in [0, 1, 5, 4]], # side face
+ [self.prism_vertices[j] for j in [1, 2, 6, 5]], # side face
+ [self.prism_vertices[j] for j in [2, 3, 7, 6]], # side face
+ [self.prism_vertices[j] for j in [3, 0, 4, 7]] # side face
+ ]
+
+ self.prism_collection = Poly3DCollection(self.prism_faces, facecolors='gray', linewidths=1, edgecolors='black',
+ alpha=0.25)
+ self.ax.add_collection3d(self.prism_collection)
+
+ # Front Arrow (Purple)
+ self.front_arrow, = self.ax.plot([0, 2], [0, 0], [0, 0], color='purple', marker='o', markersize=10,
+ label='Front Arrow')
+
+ # Up Arrow (Magenta)
+ self.up_arrow, = self.ax.plot([0, 0], [0, -1], [0, 1], color='magenta', marker='o', markersize=10,
+ label='Up Arrow')
+
+ # Side Arrow (Orange)
+ self.side_arrow, = self.ax.plot([0, 1], [0, -1], [0, 1], color='orange', marker='o', markersize=10,
+ label='Side Arrow')
+
+ # Create a canvas to draw on
+ self.canvas = FigureCanvasTkAgg(self.fig, master=self.frame1)
+ self.canvas.draw()
+ self.canvas.get_tk_widget().pack(fill=tk.BOTH, expand=True)
+
+ # Create a label for average EMG
+ # self.emg_label = tk.Label(self.frame1, text="Average EMG: 0", font=("Arial", 14))
+ # self.emg_label.pack(pady=10)
+
+ self.roll_label = tk.Label(self.frame2, text="roll is : " )
+ self.roll_label.config(font=("Arial", 12))
+ self.roll_label.place(relx=0.2, rely=0.3, anchor='w')
+ self.pitch_label = tk.Label(self.frame2, text="pitch is : " )
+ self.pitch_label.config(font=("Arial", 12))
+ self.pitch_label.place(relx=0.2, rely=0.4, anchor='w')
+ self.yaw_label = tk.Label(self.frame2, text="yaw is : " )
+ self.yaw_label.config(font=("Arial", 12))
+ self.yaw_label.place(relx=0.2, rely=0.5, anchor='w')
+
+ # initial button, commands and metrics for EMG detector
+ def _initialise_EMG_graph(self):
+ if self.input_port in self.ports:#port maybe different on different laptop
+
+ self.label2 = tk.Label(self.frame3, text=f"Port: {self.input_port} ")
+ self.label2.place(relx=0.23, rely=0.8, anchor='center')
+ self.label1 = tk.Label(self.frame3,
+ text="click the Connect button to see the animation",
+ wraplength=self.width / 2)
+ self.label1.place(relx=0.5, rely=0.9, anchor='center')
+ # Add a button to start data transmission
+ #self.start_button = tk.Button(self.frame3, text="connect", command=self.start_EMG_data_transmission)
+ #self.start_button.place(relx=0.45, rely=0.8, anchor='center')
+
+ #self.start_button = tk.Button(self.frame3, text="Disconnect", command=self.disconnect)
+ #self.start_button.place(relx=0.7, rely=0.8, anchor='center')
+
+ else:
+ print("EMG is not connected")
+ self.label2 = tk.Label(self.frame3, text="Port: None ")
+ self.label2.place(relx=0.35, rely=0.8, anchor='center')
+ self.label1 = tk.Label(self.frame3,
+ text="Please check the IUM connection",
+ wraplength=self.width / 2)
+ self.label1.place(relx=0.5, rely=0.9, anchor='center')
+
+ # Create a figure and axis
+ self.EMG_transmitting = False
+ fig = Figure(figsize=((self.width / 200), (self.height / 200))) # Adjusting figsize based on frame size
+ self.ax1 = fig.add_subplot(111)
+
+ self.ax1.set_title("Electromyography Envelope", fontsize=14, pad=0)
+
+
+ self.ax1.set_xlim(0, 5)
+ self.ax1.set_ylim(0, 5)
+
+ self.ax1.set_xlabel("Sample(20 samples per second)",fontsize=8,labelpad=-2)
+ self.ax1.set_ylabel("Magnitude",labelpad=0)
+
+ self.ax1.set_xticks(np.arange(0, 41, 8))
+ self.ax1.set_yticks(np.arange(0, 1001, 200))
+
+ for x_tick in self.ax1.get_xticks():
+ self.ax1.axvline(x_tick, color='gray', linestyle='--', linewidth=0.5)
+ for y_tick in self.ax1.get_yticks():
+ self.ax1.axhline(y_tick, color='gray', linestyle='--', linewidth=0.5)
+
+
+
+
+ # Plot two lines
+ self.line1, = self.ax1.plot([], [], color='red', label='Outer Wrist Muscle (Extensor Carpi Ulnaris)')
+ self.line2, = self.ax1.plot([], [], color='blue', label='Inner Wrist Muscle (Flexor Carpi Radialis)')
+ self.ax1.legend(fontsize=9, loc='upper right')
+
+
+
+
+ # Embed the plot in the tkinter frame
+ self.canvas1 = FigureCanvasTkAgg(fig, master=self.frame4)
+ self.canvas1.draw()
+ self.canvas1.get_tk_widget().pack(fill=tk.BOTH, expand=True)
+ self.EMG_Display()
+
+ self.outer_EMG_label = tk.Label(self.frame3, text=f"EMG for Extensor Carpi Ulnaris is :")
+ self.outer_EMG_label.config(font=("Arial", 12))
+ self.outer_EMG_label.place(relx=0.1, rely=0.2, anchor='w')
+ self.outer_EMG_Number = tk.Label(self.frame3, text="",fg="red")
+ self.outer_EMG_Number.config(font=("Arial", 12))
+ self.outer_EMG_Number.place(relx=0.2, rely=0.3, anchor='w')
+ self.inner_EMG_label = tk.Label(self.frame3, text=f"EMG for Flexor Carpi Radialis is :")
+ self.inner_EMG_label.config(font=("Arial", 12))
+ self.inner_EMG_label.place(relx=0.1, rely=0.4, anchor='w')
+ self.inner_EMG_Number = tk.Label(self.frame3, text="",fg="blue")
+ self.inner_EMG_Number.config(font=("Arial", 12))
+ self.inner_EMG_Number.place(relx=0.2, rely=0.5, anchor='w')
+ self.gesture_label = tk.Label(self.frame3, text=f"Gesture is :")
+ self.gesture_label.config(font=("Arial", 12))
+ self.gesture_label.place(relx=0.1, rely=0.6, anchor='w')
+ self.gesture_predict = tk.Label(self.frame3, text="")
+ self.gesture_predict.config(font=("Arial", 12))
+ self.gesture_predict.place(relx=0.2, rely=0.7, anchor='w')
+ self.a, self.b = self.load_Function()
+
+
+
+
+ #button effect when connect
+ def start_data_transmission(self):
+ # Set the transmitting flag to True and start the update loop
+ if self.input_port in self.ports:
+ if self.arduino==None:
+ self.arduino = serial.Serial(self.input_port, 9600)
+ self.transmitting = True
+ self.update_display()
+
+ #button effect when connect for EMG
+ def start_EMG_data_transmission(self):
+ # Set the transmitting flag to True and start the update loop
+ if self.input_port in self.ports:
+ if self.arduino==None:
+ self.arduino = serial.Serial(self.input_port, 9600)
+ self.EMG_transmitting = True
+ self.EMG_Display()
+
+ #Go to the interface for metrics
+ def game_Start(self):
+ self.root.destroy() # Close the welcome window
+ new_root = tk.Tk()
+ app = gameScreen(new_root)
+ new_root.mainloop()
+
+
+ #effect for disconnect button
+ def disconnect(self):
+ self.transmitting = False
+ self.EMG_transmitting=False
+ self.root.after_cancel(self.update_display_id)
+ if self.arduino is not None:
+ self.arduino.close()
+ self.arduino = None
+
+
+
+ #when the IMU connected show the real-time 3D metrics
+ def update_display(self):
+ if self.transmitting:
+ try:
+ while ((self.arduino.inWaiting() > 0)and
+ (self.transmitting==True)):
+ dataPacket = self.arduino.readline()
+ dataPacket = dataPacket.decode()
+ cleandata = dataPacket.replace("\r\n", "")
+ row = cleandata.strip().split(',')
+
+
+ if len(row) == 10:
+ splitPacket = cleandata.split(',')
+ q0 = float(splitPacket[2]) # qw
+ q1 = float(splitPacket[3]) # qx
+ q2 = float(splitPacket[4]) # qy
+ q3 = float(splitPacket[5]) # qz
+ emg1 = float(splitPacket[0]) # First EMG sensor data
+ emg2 = float(splitPacket[1]) # Second EMG sensor data
+ #print(f"emg1: {emg1}, emg2: {emg2}")
+ data = [emg1, emg2]
+ predictions = self.predict(data, self.a, self.b)
+ ges_predictions = None
+ if predictions is not None:
+ if predictions == -1:
+ ges_predictions = "Hand Open"
+ if predictions == 1:
+ ges_predictions = "Hand Close"
+ if predictions == 0:
+ ges_predictions = "Unknown"
+ self.gesture_predict.config(text=f"{ges_predictions}")
+
+
+ self.outer_EMG_Number.config(text=f"{emg1}")
+ self.inner_EMG_Number.config(text=f"{emg2}")
+ self.emg_data_1.append(emg1)
+ self.emg_data_1.pop(0)
+ self.emg_data_2.append(emg2)
+ self.emg_data_2.pop(0)
+
+ # Update the line data to shift the line from right to left
+ self.line1.set_data(range(len(self.emg_data_1)), self.emg_data_1)
+ self.line2.set_data(range(len(self.emg_data_2)), self.emg_data_2)
+
+ # Redraw the canvas
+ self.canvas1.draw() # Redraw the canvas
+
+ # Calculate Angles
+ roll = math.atan2(2 * (q0 * q1 + q2 * q3), 1 - 2 * (q1 * q1 + q2 * q2))
+ pitch = -math.asin(2 * (q0 * q2 - q3 * q1))
+ yaw = -math.atan2(2 * (q0 * q3 + q1 * q2), 1 - 2 * (q2 * q2 + q3 * q3))
+
+
+ self.roll_label.config( text="roll is : "+str(roll))
+ self.pitch_label.config(text="pitch is : "+str(pitch))
+ self.yaw_label.config(text="yaw is : "+str(yaw))
+ #print(roll, pitch, yaw)
+
+
+ # Rotation matrices
+ Rz = np.array([
+ [np.cos(yaw), -np.sin(yaw), 0],
+ [np.sin(yaw), np.cos(yaw), 0],
+ [0, 0, 1]
+ ])
+
+ Ry = np.array([
+ [np.cos(pitch), 0, np.sin(pitch)],
+ [0, 1, 0],
+ [-np.sin(pitch), 0, np.cos(pitch)]
+ ])
+
+ Rx = np.array([
+ [1, 0, 0],
+ [0, np.cos(roll), -np.sin(roll)],
+ [0, np.sin(roll), np.cos(roll)]
+ ])
+
+ R = Rz @ Ry @ Rx # Combined rotation matrix
+
+ # Apply the rotation
+ rotated_vertices = (R @ self.prism_vertices.T).T
+
+ prism_faces_rotated = [
+ [rotated_vertices[j] for j in [0, 1, 2, 3]], # bottom face
+ [rotated_vertices[j] for j in [4, 5, 6, 7]], # top face
+ [rotated_vertices[j] for j in [0, 1, 5, 4]], # side face
+ [rotated_vertices[j] for j in [1, 2, 6, 5]], # side face
+ [rotated_vertices[j] for j in [2, 3, 7, 6]], # side face
+ [rotated_vertices[j] for j in [3, 0, 4, 7]] # side face
+ ]
+
+ # Update the collection
+ self.prism_collection.set_verts(prism_faces_rotated)
+
+ # Update Arrows
+ k = np.array([np.cos(yaw) * np.cos(pitch), np.sin(pitch), np.sin(yaw) * np.cos(pitch)]) # X vector
+ y = np.array([0, 1, 0]) # Y vector: pointing down
+ s = np.cross(k, y) # Side vector
+ v = np.cross(s, k) # Up vector
+ vrot = v * np.cos(roll) + np.cross(k, v) * np.sin(roll) # Rotated Up vector
+
+ self.front_arrow.set_data([0, k[0] * 2], [0, k[1] * 2])
+ self.front_arrow.set_3d_properties([0, k[2] * 2])
+ self.up_arrow.set_data([0, vrot[0] * 1], [0, vrot[1] * 1])
+ self.up_arrow.set_3d_properties([0, vrot[2] * 1])
+ self.side_arrow.set_data([0, s[0] * 1], [0, s[1] * 1])
+ self.side_arrow.set_3d_properties([0, s[2] * 1])
+
+ # Update canvas
+ self.canvas.draw()
+
+
+ #self.averageemg += emg
+
+
+ # Update EMG Label
+ #self.emg_label.config(text=f"Average EMG: {self.averageemg:.2f}")
+
+ except Exception as e:
+ print(f"An error occurred: {e}")
+
+ # Call update_display() again after 50 milliseconds
+ self.update_display_id =self.root.after(1, self.update_display)
+
+ #when the EMG connected for the line metrics for EMG data
+ def EMG_Display(self):
+ data_collection_duration = 3
+ if self.EMG_transmitting:
+ try:
+ while self.arduino.inWaiting() > 0:
+ dataPacket = self.arduino.readline()
+ dataPacket = dataPacket.decode()
+ cleandata = dataPacket.replace("\r\n", "")
+ row = cleandata.strip().split(',')
+
+ if len(row) == 10:
+ splitPacket = cleandata.split(',')
+
+
+
+ except Exception as e:
+ print(f"An error occurred: {e}")
+
+ # Call update_display() again after 50 milliseconds
+ self.EMG_display_id = self.root.after(1, self.EMG_Display)
+
+
+ #decode the EMG data from detectors
+ def _decode(self, serial_data):
+ serial_string = serial_data.decode(errors="ignore")
+ adc_string_1 = ""
+ adc_string_2 = ""
+ self.adc_values = [0, 0]
+ if '\n' in serial_string:
+ # remove new line character
+ serial_string = serial_string.replace("\n", "")
+ if serial_string != '':
+ # Convert number to binary, placing 0s in empty spots
+ serial_string = format(int(serial_string, 10), "024b")
+
+ # Separate the input number from the data
+ for i0 in range(0, 12):
+ adc_string_1 += serial_string[i0]
+ for i0 in range(12, 24):
+ adc_string_2 += serial_string[i0]
+
+ self.adc_values[0] = int(adc_string_1, base=2)
+ self.adc_values[1] = int(adc_string_2, base=2)
+
+ return self.adc_values
+
+ #load trained function for predict
+ def load_Function(self,filename='trained.txt'):
+ try:
+ with open(filename, 'r') as file:
+ lines = file.readlines()
+ if len(lines) < 2:
+ raise ValueError("File content is insufficient to read the vertical line parameters.")
+
+ a = float(lines[0].strip())
+ b = float(lines[1].strip())
+ print(f"a is {a}, b is {b}")
+
+ return a,b
+
+ except FileNotFoundError:
+ raise FileNotFoundError(f"The file {filename} does not exist.")
+ except ValueError as e:
+ raise ValueError(f"Error reading the file: {e}")
+
+ #predict the gestures
+ def predict(self, point,a,b):
+ """predict whether the point is located in the left or right"""
+ x, y = point
+ line_y = a * x + b
+ if y < line_y:
+ return -1 # left
+ elif y > line_y:
+ return 1 # right
+ else:
+ return 0 # on
+
+class WelcomeWindow:
+ def __init__(self, root):
+ self.root = root
+ self.root.title("Welcome")
+ self.width = 1000
+ self.height = 600
+ screen_width = self.root.winfo_screenwidth()
+ screen_height = self.root.winfo_screenheight()
+ x = (screen_width // 2) - (self.width // 2)
+ y = (screen_height // 2) - (self.height // 2)
+ self.root.geometry(f"{self.width}x{self.height}+{x}+{y}")
+
+ # Configure the grid to be expandable
+ self.root.columnconfigure(0, weight=1)
+ self.root.columnconfigure(1, weight=1)
+ self.root.rowconfigure(0, weight=1)
+ self.root.rowconfigure(1, weight=1)
+
+ try:
+ self.bg_image = Image.open("backGrond.jpg")
+ #print("Image loaded successfully")
+ self.bg_image = self.bg_image.resize((self.width, self.height), Image.Resampling.LANCZOS)
+ self.bg_photo = ImageTk.PhotoImage(self.bg_image)
+
+ self.bg_label = tk.Label(self.root, image=self.bg_photo)
+ self.bg_label.place(x=0, y=0, relwidth=1, relheight=1)
+ except Exception as e:
+ print(f"Error loading image: {e}")
+
+ #self.frame1 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width, height=self.height)
+ #self.frame1.grid(row=0, column=0, columnspan=2, rowspan=2, sticky="nsew")
+ #self.button1 = tk.Button(self.frame1, text="Start", command=self.startButton)
+ #self.button1.place(relx=0.5, rely=0.8, anchor='center')
+ self.button1 = tk.Button(self.root, text="Start", command=self.startButton,width=18,
+ height=2, font=("Helvetica", 15))
+ self.button1.place(relx=0.8, rely=0.8, anchor='center') # Position the button relative to the root window
+
+ def startButton(self):
+ self.root.destroy() # Close the welcome window
+ new_root = tk.Tk()
+ app = trainingInterface(new_root)
+ new_root.mainloop()
+
+class trainingInterface:
+ def __init__(self, root):
+ self.input_port='COM9'
+ self.root = root
+ self.root.title("preparation Interface")
+ self.width = 1000
+ self.height = 600
+ self.width = 1000
+ self.height = 600
+ screen_width = self.root.winfo_screenwidth()
+ screen_height = self.root.winfo_screenheight()
+ x = (screen_width // 2) - (self.width // 2)
+ y = (screen_height // 2) - (self.height // 2)
+ self.root.geometry(f"{self.width}x{self.height}+{x}+{y}")
+ self.ports = [port.device for port in serial.tools.list_ports.comports()]
+
+ # Configure the grid to be expandable
+ self.root.columnconfigure(0, weight=1)
+ self.root.columnconfigure(1, weight=1)
+ self.root.rowconfigure(0, weight=1)
+ self.root.rowconfigure(1, weight=1)
+
+
+ # Create a frame
+ self.frame1 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width, height=(self.height *2/ 3))
+ self.frame1.grid(row=0, column=0, padx=10, pady=10, sticky="nsew")
+
+
+ self.frame2 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width, height=self.height *1/ 3)
+ self.frame2.grid(row=1, column=0, padx=10, pady=10, sticky="nsew")
+
+ self.initialEMGTraining()
+ if self.input_port in self.ports:
+
+ self.emg_data_1 = [-1.0] * 41
+ self.emg_data_2 = [-1.0] * 41
+ self.savingData=[]
+ self.openHandButton=tk.Button(self.frame2,text="Hand Open",command=self.EMG_connect_HandOpen,width=15, height=2,font=("Helvetica", 12))
+ self.openHandButton.place(relx=0.3, rely=0.3, anchor='center')
+ self.handCloseButton=tk.Button(self.frame2,text="Hand Close",command=self.handCloseButton,width=15, height=2,font=("Helvetica", 12))
+ self.handCloseButton.place(relx=0.7, rely=0.3, anchor='center')
+ self.gameStartButton = tk.Button(self.frame2, text="Start", command=self.startButton, width=15,
+ height=2,font=("Helvetica", 12))
+ self.gameStartButton.place(relx=0.5, rely=0.5, anchor='center')
+ if self.input_port not in self.ports:
+ self.label=tk.Label(self.frame2, text="No EMG device found, Please check the hardware connection",font=("Helvetica", 15))
+ self.label.place(relx=0.5, rely=0.3, anchor='center')
+ self.gameStartButton = tk.Button(self.frame2, text="Start", command=self.startButton, width=15,
+ height=2, font=("Helvetica", 12))
+ self.gameStartButton.place(relx=0.5, rely=0.5, anchor='center')
+
+ #Button effect for start game button
+ def startButton(self):
+ self.root.destroy() # Close the welcome window
+ new_root = tk.Tk()
+ app = Window(new_root)
+ new_root.mainloop()
+
+ # Button effect for start recording when the gesture is handOpen
+ def EMG_connect_HandOpen(self):
+ self.arduino_EMG = serial.Serial(self.input_port, 9600, timeout=1)
+ gesture = "handOpen"
+ self.start_countdown(11)
+ self.displayAndsaveDate()
+
+ # Button effect for start recording when the gesture is handClose
+ def handCloseButton(self):
+ self.arduino_EMG = serial.Serial(self.input_port, 9600, timeout=1)
+ gesture = "handOpen"
+ self.start_countdown_close(11)
+ self.displayAndsaveDate()
+
+ # Button effect for disconnect
+ def EMG_disconnect(self):
+ if self.arduino_EMG is not None:
+ self.arduino_EMG.close()
+ self.arduino_EMG = None
+
+ #show the countdown numbers when record hand open
+ def start_countdown(self, count):
+ if count > 0:
+ self.startSave=True
+ if count<11:
+ self.openHandButton.config(text=str(count))
+ self.frame2.after(1000, self.start_countdown, count - 1)
+ else:
+ self.openHandButton.config(text="Hand Open")
+ self.startSave = False
+ self.savedDataOpen = []
+ for i in self.savingData:
+ self.savedDataOpen.append(i)
+ print(f"open: {self.savedDataOpen}")
+ self.savingData.clear()
+ self.EMG_disconnect()
+
+ #show the countdown numbers hwne record hand close
+ def start_countdown_close(self, count):
+ if count > 0:
+ self.startSave=True
+ if count<11:
+ self.handCloseButton.config(text=str(count))
+ self.frame2.after(1000, self.start_countdown_close, count - 1)
+ else:
+ self.handCloseButton.config(text="Hand Close")
+ self.startSave = False
+ self.savedDataClose=[]
+ for i in self.savingData:
+ self.savedDataClose.append(i)
+ self.savingData.clear()
+ print(f"close:{self.savedDataClose}")
+ self.EMG_disconnect()
+ self.trainData()
+
+
+ #save data
+ def displayAndsaveDate(self):
+ if self.startSave:
+ try:
+ while ((self.arduino_EMG.inWaiting() > 0) ):
+ dataPacket = self.arduino_EMG.readline()
+ dataPacket = dataPacket.decode()
+ cleandata = dataPacket.replace("\r\n", "")
+ row = cleandata.strip().split(',')
+
+ if len(row) == 10:
+ splitPacket = cleandata.split(',')
+ emg1 = float(splitPacket[0]) # First EMG sensor data
+ emg2 = float(splitPacket[1]) # Second EMG sensor data
+ print(f"emg1: {emg1}, emg2: {emg2}")
+
+
+ self.emg_data_1.append(emg1)
+ self.emg_data_1.pop(0)
+ self.emg_data_2.append(emg2)
+ self.emg_data_2.pop(0)
+ if self.startSave==True:
+ self.savingData.append([emg1,emg2])
+ print(len(self.savingData))
+
+
+ # Update the line data to shift the line from right to left
+ self.line1.set_data(range(len(self.emg_data_1)), self.emg_data_1)
+ self.line2.set_data(range(len(self.emg_data_2)), self.emg_data_2)
+
+ # Redraw the canvas
+ self.canvas1.draw() # Redraw the canvas
+
+ except Exception as e:
+ print(f"An error occurred: {e}")
+
+ self.EMG_display_id = self.root.after(1, self.displayAndsaveDate)
+
+
+
+ #draw the buttons, metrics for initial training interface
+ def initialEMGTraining(self):
+ self.EMG_transmitting = False
+ fig = Figure(figsize=(self.frame1.winfo_width() / 100, self.frame1.winfo_height() / 100))
+ self.ax1 = fig.add_subplot(111)
+
+ self.ax1.set_title("Electromyography Envelope", fontsize=14, pad=0)
+ self.ax1.set_xlim(0, 5)
+ self.ax1.set_ylim(0, 5)
+ self.ax1.set_xlabel("Sample (20 samples per second)", fontsize=8, labelpad=-2)
+ self.ax1.set_ylabel("Magnitude", labelpad=0)
+ self.ax1.set_xticks(np.arange(0, 41, 8))
+ self.ax1.set_yticks(np.arange(0, 1001, 200))
+
+ for x_tick in self.ax1.get_xticks():
+ self.ax1.axvline(x_tick, color='gray', linestyle='--', linewidth=0.5)
+ for y_tick in self.ax1.get_yticks():
+ self.ax1.axhline(y_tick, color='gray', linestyle='--', linewidth=0.5)
+
+ self.line1, = self.ax1.plot([], [], color='red', label='Outer Wrist Muscle (Extensor Carpi Ulnaris)')
+ self.line2, = self.ax1.plot([], [], color='blue', label='Inner Wrist Muscle (Flexor Carpi Radialis)')
+ self.ax1.legend(fontsize=9, loc='upper right')
+
+ # Embed the plot in the tkinter frame
+ self.canvas1 = FigureCanvasTkAgg(fig, master=self.frame1)
+ self.canvas1.draw()
+ self.canvas1.get_tk_widget().pack(fill=tk.BOTH, expand=True)
+
+ # Bind the resizing event to the figure update
+ self.frame1.bind("<Configure>", self.on_frame_resize)
+
+ def on_frame_resize(self, event):
+ width = self.frame1.winfo_width()
+ height = self.frame1.winfo_height()
+ self.canvas1.get_tk_widget().config(width=width, height=height)
+ self.canvas1.draw()
+
+ '''
+ Train Data
+ '''
+
+ #turn the saved data to the algorithm and save to the file trained.txt
+ def trainData(self):
+ if os.path.exists('trained.txt'):
+ os.remove('trained.txt')
+
+ if (self.savedDataClose != []) and (self.savedDataClose != []):
+ data, labels=self.prepare_data(self.savedDataClose,self.savedDataClose)
+ vertical_line = Algorithm(self.savedDataClose, self.savedDataOpen)
+ print(f"Function is: y = {vertical_line.a}x + {vertical_line.b}")
+
+ with open('trained.txt', 'w') as file:
+ file.write(f"{vertical_line.a}\n")
+ file.write(f"{vertical_line.b}\n")
+
+ return vertical_line
+
+ def _decode(self, serial_data):
+ serial_string = serial_data.decode(errors="ignore")
+ adc_string_1 = ""
+ adc_string_2 = ""
+ self.adc_values = [0, 0]
+ if '\n' in serial_string:
+ # remove new line character
+ serial_string = serial_string.replace("\n", "")
+ if serial_string != '':
+ # Convert number to binary, placing 0s in empty spots
+ serial_string = format(int(serial_string, 10), "024b")
+
+ # Separate the input number from the data
+ for i0 in range(0, 12):
+ adc_string_1 += serial_string[i0]
+ for i0 in range(12, 24):
+ adc_string_2 += serial_string[i0]
+
+ self.adc_values[0] = int(adc_string_1, base=2)
+ self.adc_values[1] = int(adc_string_2, base=2)
+
+ return self.adc_values
+
+ def prepare_data(self,list1, list2):
+ data = torch.tensor(list1 + list2, dtype=torch.float32)
+ labels = torch.cat((torch.zeros(len(list1)), torch.ones(len(list2))), dim=0)
+ return data, labels
+
+
+class Algorithm:
+ def __init__(self, list1, list2):
+ self.a, self.b = self.calculate_line_equation(list1, list2)
+
+ def calculate_average(self, lst):
+ """calculate the average of these points"""
+ n = len(lst)
+ if n == 0:
+ return (0, 0)
+ sum_x = sum(point[0] for point in lst)
+ sum_y = sum(point[1] for point in lst)
+ return (sum_x / n, sum_y / n)
+
+ def calculate_line_equation(self, list1, list2):
+ """calculate the line equation in the form of y = ax + b"""
+ avg1 = self.calculate_average(list1)
+ avg2 = self.calculate_average(list2)
+
+ x1, y1 = avg1
+ x2, y2 = avg2
+
+ # Calculating the slope
+ if x1 == x2:
+ raise ValueError("The slope of a vertical line is undefined because two points are on the same vertical line.")
+
+ slope = (y2 - y1) / (x2 - x1)
+
+ perpendicular_slope = -1 / slope
+
+ # Use the point-slope form to convert the equation y - y1 = m(x - x1) to the form y = ax + b
+ a = perpendicular_slope
+ b = y1 - a * x1
+
+ return a, b
+
+ def predict(self, point):
+ x, y = point
+ line_y = self.a * x + self.b
+ if y < line_y:
+ return -1 # left
+ elif y > line_y:
+ return 1 # right
+ else:
+ return 0 # on
+
+class gameScreen:
+ def __init__(self, root):
+ self.input_port='COM9'
+ self.root = root
+ self.root.title("game Interface")
+ self.width = 1000
+ self.height = 600
+ self.width = 1000
+ self.height = 600
+ screen_width = self.root.winfo_screenwidth()
+ screen_height = self.root.winfo_screenheight()
+ x = (screen_width // 2) - (self.width // 2)
+ y = (screen_height // 2) - (self.height // 2)
+ self.root.geometry(f"{self.width}x{self.height}+{x}+{y}")
+ self.ports = [port.device for port in serial.tools.list_ports.comports()]
+
+ # Configure the grid to be expandable
+ self.root.columnconfigure(0, weight=1)
+ self.root.columnconfigure(1, weight=1)
+ self.root.rowconfigure(0, weight=1)
+ self.root.rowconfigure(1, weight=1)
+
+ # Create a frame
+ self.frame1 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width, height=(self.height * 1 / 2))
+ self.frame1.grid(row=0, column=0, padx=10, pady=10, sticky="nsew")
+
+ self.frame2 = tk.Frame(self.root, borderwidth=1, relief="solid", width=self.width, height=self.height * 1 / 2)
+ self.frame2.grid(row=1, column=0, padx=10, pady=10, sticky="nsew")
+
+ if self.input_port in self.ports :
+ #self.arduino_EMG = serial.Serial(self.input_port, 9600, timeout=1)
+ self.outer_EMG_label = tk.Label(self.frame2, text=f"EMG for Extensor Carpi Ulnaris is :")
+ self.outer_EMG_label.config(font=("Arial", 12))
+ self.outer_EMG_label.place(relx=0.1, rely=0.2, anchor='w')
+ self.outer_EMG_Number = tk.Label(self.frame2, text="", fg="red")
+ self.outer_EMG_Number.config(font=("Arial", 12))
+ self.outer_EMG_Number.place(relx=0.2, rely=0.3, anchor='w')
+ self.inner_EMG_label = tk.Label(self.frame2, text=f"EMG for Flexor Carpi Radialis is :")
+ self.inner_EMG_label.config(font=("Arial", 12))
+ self.inner_EMG_label.place(relx=0.1, rely=0.4, anchor='w')
+ self.inner_EMG_Number = tk.Label(self.frame2, text="", fg="blue")
+ self.inner_EMG_Number.config(font=("Arial", 12))
+ self.inner_EMG_Number.place(relx=0.2, rely=0.5, anchor='w')
+ self.gesture_label = tk.Label(self.frame2, text=f"Gesture is :")
+ self.gesture_label.config(font=("Arial", 12))
+ self.gesture_label.place(relx=0.1, rely=0.6, anchor='w')
+ self.gesture_predict = tk.Label(self.frame2, text="")
+ self.gesture_predict.config(font=("Arial", 12))
+ self.gesture_predict.place(relx=0.2, rely=0.7, anchor='w')
+ self.a, self.b = self.load_Function()
+ #self.emg_thread = threading.Thread(target=self.EMG_Display)
+ # self.emg_thread.start()
+
+
+ #self.EMG_Display()
+ if self.input_port in self.ports:
+ self.column_limit = 9
+ self.last_averageRoll = 0
+ self.last_averageyaw = 0
+ self.last_averagePitch = 0
+
+ self.averageroll = 0
+ self.averageyaw = 0
+ self.averagepitch = 0
+ self.last_print_time = time()
+ self.arduino = serial.Serial(self.input_port, 115200)
+ self.roll_label = tk.Label(self.frame1, text="roll is : ")
+ self.roll_label.config(font=("Arial", 12))
+ self.roll_label.place(relx=0.2, rely=0.3, anchor='w')
+ self.pitch_label = tk.Label(self.frame1, text="pitch is : ")
+ self.pitch_label.config(font=("Arial", 12))
+ self.pitch_label.place(relx=0.2, rely=0.4, anchor='w')
+ self.yaw_label = tk.Label(self.frame1, text="yaw is : ")
+ self.yaw_label.config(font=("Arial", 12))
+ self.yaw_label.place(relx=0.2, rely=0.5, anchor='w')
+ self.IMU_Display()
+ #self.imu_thread = threading.Thread(target=self.IMU_Display)
+ #self.imu_thread.start()
+ #self.IMU_Display()
+
+
+ def _decode(self, serial_data):
+ serial_string = serial_data.decode(errors="ignore")
+ adc_string_1 = ""
+ adc_string_2 = ""
+ self.adc_values = [0, 0]
+ if '\n' in serial_string:
+ # remove new line character
+ serial_string = serial_string.replace("\n", "")
+ if serial_string != '':
+ # Convert number to binary, placing 0s in empty spots
+ serial_string = format(int(serial_string, 10), "024b")
+
+ # Separate the input number from the data
+ for i0 in range(0, 12):
+ adc_string_1 += serial_string[i0]
+ for i0 in range(12, 24):
+ adc_string_2 += serial_string[i0]
+
+ self.adc_values[0] = int(adc_string_1, base=2)
+ self.adc_values[1] = int(adc_string_2, base=2)
+
+ return self.adc_values
+
+ #display the real-time IMU data and sent to Unity by Socket
+ def IMU_Display(self):
+ try:
+ while (self.arduino.inWaiting() > 0):
+ dataPacket = self.arduino.readline()
+ dataPacket = dataPacket.decode()
+ cleandata = dataPacket.replace("\r\n", "")
+ row = cleandata.strip().split(',')
+
+
+ if len(row) == 10:
+ splitPacket = cleandata.split(',')
+ q0 = float(splitPacket[2]) # qw
+ q1 = float(splitPacket[3]) # qx
+ q2 = float(splitPacket[4]) # qy
+ q3 = float(splitPacket[5]) # qz
+ emg1 = float(splitPacket[0]) # First EMG sensor data
+ emg2 = float(splitPacket[1]) # Second EMG sensor data
+ #print(f"emg1: {emg1}, emg2: {emg2}")
+ data = [emg1, emg2]
+ predictions = self.predict(data, self.a, self.b)
+ ges_predictions = None
+ if predictions is not None:
+ if predictions == -1:
+ ges_predictions = "Hand Open"
+ if predictions == 1:
+ ges_predictions = "Hand Close"
+ if predictions == 0:
+ ges_predictions = "Unknown"
+ self.gesture_predict.config(text=f"{ges_predictions}")
+ self.outer_EMG_Number.config(text=f"{emg1}")
+ self.inner_EMG_Number.config(text=f"{emg2}")
+ self.send_command_to_unity(f"Hand :{ges_predictions}")
+
+
+ # Calculate Angles
+ roll = math.atan2(2 * (q0 * q1 + q2 * q3), 1 - 2 * (q1 * q1 + q2 * q2))
+ pitch = -math.asin(2 * (q0 * q2 - q3 * q1))
+ yaw = -math.atan2(2 * (q0 * q3 + q1 * q2), 1 - 2 * (q2 * q2 + q3 * q3))
+
+
+ self.roll_label.config( text="roll is : "+str(roll))
+ self.pitch_label.config(text="pitch is : "+str(pitch))
+ self.yaw_label.config(text="yaw is : "+str(yaw))
+ print(roll, pitch, yaw)
+ current_time = time()
+ if current_time - self.last_print_time >= 0.01:
+ print(f"roll is: {roll}")
+ print(f"last roll is: {self.last_averageRoll}")
+ differ_roll = self.last_averageRoll - roll
+ print(f"differ roll is: {differ_roll}")
+ CalculatedAngle = differ_roll * 3000 / 2.5
+ print(f"CalculatedAngle is: {CalculatedAngle}")
+ if (differ_roll) > 0:
+ print("send_command_to_unity"+(f"Command : down {CalculatedAngle}"))
+ self.send_command_to_unity(f"Command : down {CalculatedAngle}")
+ if (differ_roll) < 0:
+ print("send_command_to_unity" + (f"Command : down {CalculatedAngle}"))
+ self.send_command_to_unity(f"Command : up {-CalculatedAngle}")
+
+ if (yaw < 0):
+ yaw = -yaw
+
+ print(f"yaw is: {yaw}")
+ print(f"last yaw is: {self.last_averageyaw}")
+ differ_yaw = self.last_averageyaw - yaw
+ print(f"differ yaw is: {differ_yaw}")
+ yawAngle = differ_yaw * 90 / 2
+ print(f"yawAngle is: {yawAngle}")
+ if (differ_yaw) < 0:
+ print("send_command_to_unity"+(f"Command : back {-yawAngle}"))
+ self.send_command_to_unity(f"Command : back {-yawAngle}")
+ if (differ_yaw) > 0:
+ print("send_command_to_unity" + (f"Command : roll {-yawAngle}"))
+ self.send_command_to_unity(f"Command : roll {yawAngle}")
+
+ self.last_print_time = current_time
+ self.last_averageRoll = roll
+ self.last_averageyaw = yaw
+ self.last_averagePitch = pitch
+
+
+
+ except Exception as e:
+ print(f"An error occurred: {e}")
+
+ # Call update_display() again after 50 milliseconds
+ self.update_display_id =self.root.after(1, self.IMU_Display)
+
+ #load the function for predict
+ def load_Function(self,filename='trained.txt'):
+ try:
+ with open(filename, 'r') as file:
+ lines = file.readlines()
+ if len(lines) < 2:
+ raise ValueError("File content is insufficient to read the vertical line parameters.")
+
+ a = float(lines[0].strip())
+ b = float(lines[1].strip())
+ print(f"a is {a}, b is {b}")
+
+ return a,b
+
+ except FileNotFoundError:
+ raise FileNotFoundError(f"The file {filename} does not exist.")
+ except ValueError as e:
+ raise ValueError(f"Error reading the file: {e}")
+
+ def predict(self, point, a, b):
+ x, y = point
+ line_y = a * x + b
+ if y < line_y:
+ return -1 # left
+ elif y > line_y:
+ return 1 # right
+ else:
+ return 0 # on
+
+ #build the connect to unity with Socket and send the response command
+ def send_command_to_unity(self,command):
+ host = '127.0.0.1' # IP address for the Unity server
+ port = 65432 # The port that the Unity server listens on
+
+ with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+ s.connect((host, port))
+ s.sendall(command.encode())
+ response = s.recv(1024)
+ print('Received', repr(response))
+
+
+
+
+
+if __name__ == "__main__":
+ root1 = tk.Tk()
+ appWelcome = WelcomeWindow(root1)
+ root1.mainloop()
diff --git a/integration/Window2.py b/integration/Window2.py
index dec807ce2f37cc1061613fe520d17d5d3ca62054..7eb0fb177ab2d6da810ba99181d23d4b947322b7 100644
--- a/integration/Window2.py
+++ b/integration/Window2.py
@@ -14,6 +14,8 @@ import os
from PIL import Image, ImageTk
from time import sleep, time
+#this is the file for the combined hardware detector
+
class Window:
def __init__(self, root):
self.input_port='COM9'
diff --git a/integration/mlp.py b/integration/mlp.py
new file mode 100644
index 0000000000000000000000000000000000000000..fb43f9e85e7ab7739583ee5ffe610914ceb321b8
--- /dev/null
+++ b/integration/mlp.py
@@ -0,0 +1,64 @@
+import torch
+import torch.nn as nn
+import torch.optim as optim
+from sklearn.model_selection import train_test_split
+
+# 生成一些随机数据
+def generate_data(num_samples):
+ # 生成两类数据
+ label_0 = torch.randn(num_samples, 2) + torch.tensor([1, 1])
+ label_1 = torch.randn(num_samples, 2) + torch.tensor([-1, -1])
+ labels = torch.cat((torch.zeros(num_samples), torch.ones(num_samples)), dim=0)
+ data = torch.cat((label_0, label_1), dim=0)
+ return data, labels
+
+# 定义MLP模型
+class MLP(nn.Module):
+ def __init__(self):
+ super(MLP, self).__init__()
+ self.fc1 = nn.Linear(2, 16)
+ self.fc2 = nn.Linear(16, 16)
+ self.fc3 = nn.Linear(16, 2) # 二分类问题
+
+ def forward(self, x):
+ x = torch.relu(self.fc1(x))
+ x = torch.relu(self.fc2(x))
+ x = self.fc3(x)
+ return x
+
+# 超参数
+num_samples = 1000
+num_epochs = 50
+batch_size = 32
+learning_rate = 0.01
+
+# 数据准备
+data, labels = generate_data(num_samples)
+train_data, test_data, train_labels, test_labels = train_test_split(data, labels, test_size=0.2)
+
+# 数据集和数据加载器
+train_dataset = torch.utils.data.TensorDataset(train_data, train_labels)
+train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
+
+# 初始化模型、损失函数和优化器
+model = MLP()
+criterion = nn.CrossEntropyLoss()
+optimizer = optim.Adam(model.parameters(), lr=learning_rate)
+
+# 训练模型
+for epoch in range(num_epochs):
+ for inputs, targets in train_loader:
+ optimizer.zero_grad()
+ outputs = model(inputs)
+ loss = criterion(outputs, targets.long())
+ loss.backward()
+ optimizer.step()
+ print(f'Epoch [{epoch + 1}/{num_epochs}], Loss: {loss.item():.4f}')
+
+# 测试模型
+model.eval()
+with torch.no_grad():
+ test_outputs = model(test_data)
+ _, predicted = torch.max(test_outputs, 1)
+ accuracy = (predicted == test_labels).sum().item() / len(test_labels)
+ print(f'Accuracy on test data: {accuracy * 100:.2f}%')