Subindo arquivos

README.md

@@ -1,5 +1,5 @@
 ---
-title: Gtm Keras H5 Predictor
+title: Gtm-Keras-H5-Predictor
 emoji: 😻
 colorFrom: green
 colorTo: red
@@ -10,4 +10,22 @@ pinned: false
 license: ecl-2.0
 ---
 
+## Description
+In Google Teachable Machine, after training, under 'Export Model', go to 'Tensorflow', click on 'Keras' and then 'Download my model' (wait a moment). The zip will contain the Keras .h5 model. This application allows users to upload these models and test them with their own images.
+
+## How to Use
+1. Train a model in Google Teachable Machine.
+2. Export and download the model as a Keras .h5 file (in Tensorflow, in middle, click on Keras, Download my model).
+3. Upload the model (.zip or .h5) and your test images to the application.
+4. View the predictions made by the model.
+
+## Developer
+Developed by Ramon Mayor Martins (2023)
+
+- E-mail: [rmayormartins@gmail.com](mailto:rmayormartins@gmail.com)
+- Homepage: [https://rmayormartins.github.io/](https://rmayormartins.github.io/)
+- Twitter: [@rmayormartins](https://twitter.com/rmayormartins)
+- GitHub: [https://github.com/rmayormartins](https://github.com/rmayormartins)
+
+
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,56 @@
+import gradio as gr
+from tensorflow.keras.models import load_model
+from PIL import Image
+import numpy as np
+import zipfile
+import tempfile
+import os
+
+def classify_images(uploaded_file, images):
+    
+    with tempfile.TemporaryDirectory() as temp_dir:
+        with zipfile.ZipFile(uploaded_file, 'r') as zip_ref:
+            zip_ref.extractall(temp_dir)
+        model_path = os.path.join(temp_dir, 'keras_model.h5')
+        labels_path = os.path.join(temp_dir, 'labels.txt')
+
+        
+        loaded_model = load_model(model_path)
+
+        
+        with open(labels_path, 'r') as file:
+            class_names = file.read().splitlines()
+
+        
+        predictions = []
+        for img in images:
+            with Image.open(img) as pil_image:
+                image = pil_image.resize((224, 224))
+                image = np.array(image)
+                image = image / 255.0
+                image = np.expand_dims(image, axis=0)
+
+                prediction = loaded_model.predict(image)
+                predicted_class = class_names[np.argmax(prediction)]
+                predictions.append(predicted_class)
+
+        return predictions
+
+iface = gr.Interface(
+    fn=classify_images,
+    inputs=[
+        gr.File(label="Upload do Modelo (.h5 or .zip (with .h5))"),
+        gr.Files(label="Upload of Images")
+    ],
+    outputs="text",
+    title="GTM-Keras-h5-Predictor",
+    description="In Google Teachable Machine, after training, under 'Export Model', go to 'Tensorflow', click on 'Keras' and then 'Download my model' (wait a moment). The zip will contain the Keras .h5 model."
+)
+
+iface.add_example(inputs=["converted_keras.zip", ["example1.jpg", "example2.jpg"]])
+
+
+if __name__ == "__main__":
+    iface.launch(debug=True)
+
+

converted_keras.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac6edc498947ade81a3d42a6a4140f55c9e5e8baccba06429c1ba4c798132a0a
+size 2453667

requirements.txt

@@ -0,0 +1,4 @@
+gradio
+tensorflow
+Pillow
+numpy