komit

app.py

@@ -1,7 +1,49 @@
-import gradio as gr
+import os
+import torch
+import torchvision.transforms as transforms
+from torch.utils.data import DataLoader
+from PIL import Image
+from models import ResNet18
+from datasets import HandGestureDataset
 
-def greet(name):
-    return "Hello " + name + "!!"
+# Set the path to the dataset directory
+data_dir = 'dataset'
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch()
+# Define the image transforms
+transform = transforms.Compose([
+    transforms.Resize((224, 224)),
+    transforms.ToTensor(),
+    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
+])
+
+# Define the dataset
+dataset = HandGestureDataset(data_dir, transform=transform)
+
+# Define the data loader
+dataloader = DataLoader(dataset, batch_size=32, shuffle=True)
+
+# Load the pre-trained neural network
+model = ResNet18(pretrained=True)
+
+# Replace the final fully connected layer with a new one
+num_classes = 7
+model.fc = torch.nn.Linear(model.fc.in_features, num_classes)
+
+# Set the model to evaluation mode
+model.eval()
+
+# Load an image from the dataset
+for i, (image, label) in enumerate(dataloader):
+    # Apply the image transforms
+    image = transform(image)
+
+    # Add a batch dimension
+    image = image.unsqueeze(0)
+
+    # Make a prediction on the image
+    with torch.no_grad():
+        output = model(image)
+        prediction = torch.argmax(output)
+
+    # Print the prediction
+    print(f'Image {i+1}, Predicted note: {prediction.item()}')

datasets.py

@@ -0,0 +1,28 @@
+import os
+import torch
+from torchvision.transforms import transforms
+from PIL import Image
+
+class HandGestureDataset(torch.utils.data.Dataset):
+    def __init__(self, data_dir, transform=None):
+        self.data_dir = data_dir
+        self.transform = transform
+        self.image_files = [os.path.join(self.data_dir, f) for f in os.listdir(self.data_dir) if f.endswith('.jpg')]
+
+    def __len__(self):
+        return len(self.image_files)
+
+    def __getitem__(self, idx):
+        image_path = self.image_files[idx]
+        image = Image.open(image_path)
+
+        if self.transform:
+            image = self.transform(image)
+
+        label = self.get_label(image_path)
+
+        return image, label
+
+    def get_label(self, image_path):
+        label = os.path.basename(os.path.dirname(image_path))
+        return label

download_image.py

@@ -0,0 +1,42 @@
+import os
+import requests
+
+def download_images(url_list, output_folder):
+    # Kreiranje izlaznog foldera ako ne postoji
+    os.makedirs(output_folder, exist_ok=True)
+
+    # Iteriranje kroz listu URL-ova
+    for i, url in enumerate(url_list):
+        try:
+            # Preuzimanje slike sa URL-a
+            response = requests.get(url)
+            if response.status_code == 200:
+                # Extrakcija naziva datoteke iz URL-a
+                filename = url.split("/")[-1]
+                # Razdvajanje naziva datoteke na osnovu nota
+                note = filename.split("_")[0]
+                # Kreiranje podfoldera za notu ako ne postoji
+                note_folder = os.path.join(output_folder, note)
+                os.makedirs(note_folder, exist_ok=True)
+                # Čuvanje slike u odgovarajućem podfolderu
+                with open(os.path.join(note_folder, filename), "wb") as f:
+                    f.write(response.content)
+                print(f"Slika {i+1} uspešno preuzeta i sačuvana u {note_folder}.")
+            else:
+                print(f"Greska prilikom preuzimanja slike {i+1}. Status kod: {response.status_code}")
+        except Exception as e:
+            print(f"Greska prilikom preuzimanja slike {i+1}: {str(e)}")
+
+# Lista URL-ova sa kojih ćemo preuzimati slike (primer)
+url_list = [
+    "https://example.com/do_image1.jpg",
+    "https://example.com/re_image1.jpg",
+    "https://example.com/mi_image1.jpg",
+    # Dodajte ostale URL-ove prema potrebi
+]
+
+# Folder u koji ćemo sačuvati preuzete slike
+output_folder = "dataset"
+
+# Pozivanje funkcije za preuzimanje slika
+download_images(url_list, output_folder)

models.py

@@ -0,0 +1,22 @@
+import torch
+import torch.nn as nn
+
+class Net(nn.Module):
+    def __init__(self):
+        super(Net, self).__init__()
+        self.conv1 = nn.Conv2d(3, 16, 3, padding=1)
+        self.conv2 = nn.Conv2d(16, 32, 3, padding=1)
+        self.conv3 = nn.Conv2d(32, 64, 3, padding=1)
+        self.pool = nn.MaxPool2d(2, 2)
+        self.fc1 = nn.Linear(64 * 28 * 28, 512)
+        self.fc2 = nn.Linear(512, 7)
+        self.dropout = nn.Dropout(0.2)
+
+    def forward(self, x):
+        x = self.pool(F.relu(self.conv1(x)))
+        x = self.pool(F.relu(self.conv2(x)))
+        x = self.pool(F.relu(self.conv3(x)))
+        x = x.view(-1, 64 * 28 * 28)
+        x = self.dropout(F.relu(self.fc1(x)))
+        x = self.fc2(x) 
+        return x