Upload 28 files

.gitignore

@@ -0,0 +1,7 @@
+Dataset/training/Cat
+Dataset/training/Dog
+Dataset/validation/Cat
+Dataset/validation/Dog
+
+venv
+rename.sh

README.md

@@ -1,13 +1,55 @@
----
-title: Dogs Cats
-emoji: 🌍
-colorFrom: green
-colorTo: green
-sdk: gradio
-sdk_version: 3.12.0
-app_file: app.py
-pinned: false
-license: openrail
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Modelo Clasificatorio de perros y gatos
+
+Nuestro proyecto es un modelo que clasifica perros y gatos, fue entrenado y validado a partir de imágenes y fue modelado con redes neuronales convolucionales.
+
+## Decisions
+
+`1` Utilizar redes neuronales convolucionales, ya que estas son capaces de clasificar de forma mas precisa imágenes a diferencia de una neuronal, y son las indicadas para los casos clasificatorios a diferencia de un modelo denso que se usa para regresiones y no es capaz de acertar en la mayoria de los casos.
+
+`2` Clasificar perros y gatos.
+
+`3` Añadir un dropout en las capas para que el modelo no se sobreajuste.
+
+`4` Anadir capas de pooling para agrupar las decisiones tomadas entre cada capa convolucional.
+
+`5` Utilizar la funcion de activacion relu que es de las mas populares por sus buenos resultados en casos de clasificacion.
+
+`6` Utilizar la funcion de activacion sigmoid en al ultima capa para un mejor resultado final.
+
+`7` Utilizar optimizador adam y funcion de perdida binary_crossentropy que tambien son de las mas recomendadas y usadas para estos casos.
+
+`8` Utilizar transformaciones en algunas imágenes de entrenamiento para evitar que el modelo se ajuste y luego no pueda clasificar imágenes con mas zoom, o que esten rotadas, etc.
+
+`9` Utilizar un tamaño para las imágenes de 100x100 para disminuir el tiempo de entrenamiento ya que con un tamaño de 400x400 inicialmente nos tomo aproximadamente una hora.
+
+`10` Utilizar 1800 imágenes de entrenamiento y 600 imágenes de validacion.
+
+## Data Sources
+
+Usamos las imágenes de entrenamiento y de validacion sobre gatos y perros de este link provenientes de TensorFlow [Cat vs Dog](https://download.microsoft.com/download/3/E/1/3E1C3F21-ECDB-4869-8368-6DEBA77B919F/kagglecatsanddogs_5340.zip).
+
+## Features
+
+La entrada pueden ser con imágenes o con el animal directamente, a traves de la camara del computador en este caso, para predecir y/o clasificar si es un gato o un perro.
+
+## Data Collection
+
+Decidimos utilizar esta coleccion de datos porque tiene gran variedad de imágenes, donde no solo sale un gato o perro en la misma forma, sino que con diferentes elementos acompañando, la imágenes son de diferentes tamaños, resoluciones, orientaciones, hay algunas con multiples gatos/perros, tienen texto, son carteles, etc.
+
+## Value Proposition
+
+Puede ser utilizado en aplicaciones de reconocimiento y distincion de estos animales, en programas de seleccion, en aplicaciones de competencias de gatos o perros para permitir la inscripcion, en juegos educativos, en aplicaciones de imagenes para categorizar, etc.
+
+# Environment requirements to run the model
+
+Python 3.8.0
+
+Tensorflow 2.2.2
+
+Numpy 1.19.4
+
+Scipy 1.9.3
+
+Matplotlib.pyplot
+
+Cv2

logs/cnn/train/events.out.tfevents.1670589111.DESKTOP-DHTLKO8.21804.243.v2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04a90b08af638d4fab1df8e8b62a43f87f29ca2f08e1909bfee8ba824a8f82d3
+size 29172

logs/cnn/train/events.out.tfevents.1670589111.DESKTOP-DHTLKO8.profile-empty

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1ad8c819b06bc5c55f396bfa486d227679ac3f1b9db4b283da509f52174df93
+size 40

logs/cnn/train/events.out.tfevents.1670589908.DESKTOP-DHTLKO8.4388.243.v2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad4206d06b35bfb6ea3e5ae38442e7183983f76c453991dead519223e8210826
+size 19856

logs/cnn/train/events.out.tfevents.1670590244.DESKTOP-DHTLKO8.29980.243.v2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9dddaa362bcd60fb8f8d95be8ba1fbf07a25194ad158f8a1a086cfd9406a09c
+size 66300

logs/cnn/train/plugins/profile/2022_12_09_12_31_51/DESKTOP-DHTLKO8.input_pipeline.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95bce5f2cf2e3a268edd53ea5a995a880068a9b79a26c18d1eeab352b1847719
+size 2077

logs/cnn/train/plugins/profile/2022_12_09_12_31_51/DESKTOP-DHTLKO8.kernel_stats.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
+size 0

logs/cnn/train/plugins/profile/2022_12_09_12_31_51/DESKTOP-DHTLKO8.overview_page.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cd4e4d6e16eb6a8cc41f8bc1e0cc503d711a401e197bf1e9371224341f0c2fc
+size 3011

logs/cnn/train/plugins/profile/2022_12_09_12_31_51/DESKTOP-DHTLKO8.tensorflow_stats.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:464d368bf7332b5706b2c8d5b2bebc00b2c4e2355042c25e912e52d03fc8da4e
+size 36508

logs/cnn/train/plugins/profile/2022_12_09_12_31_51/DESKTOP-DHTLKO8.trace.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff487b0f4b4b9c6a9b4b544b61e40a291dfd2f374b0c695d93c93dd7dbcb1896
+size 4792

logs/cnn/train/plugins/profile/2022_12_09_12_45_09/DESKTOP-DHTLKO8.input_pipeline.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df49a55d2e9da208c7136495bc7fee2095a6ad88d8393d90fdf5560c6e8c5870
+size 2077

logs/cnn/train/plugins/profile/2022_12_09_12_45_09/DESKTOP-DHTLKO8.kernel_stats.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
+size 0

logs/cnn/train/plugins/profile/2022_12_09_12_45_09/DESKTOP-DHTLKO8.overview_page.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf2f60377ddb1dc1f65b5470a804953f94f553237ae8c4e7edbb027b493058c7
+size 3011

logs/cnn/train/plugins/profile/2022_12_09_12_45_09/DESKTOP-DHTLKO8.tensorflow_stats.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02b7c35f63888a511586ff2eacafa77728b983d9e7ea5a78034ceb0a5e0ceb4f
+size 36508

logs/cnn/train/plugins/profile/2022_12_09_12_45_09/DESKTOP-DHTLKO8.trace.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6211d0462ed5c8e1f0686da20e86763779b66f9dd298e2aa38d8e8424036ab9e
+size 4854

logs/cnn/train/plugins/profile/2022_12_09_12_50_45/DESKTOP-DHTLKO8.input_pipeline.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61bb706e800ffb05421c1c4045365ec2b5b694cc9560b904de0c6cf18091da73
+size 2077

logs/cnn/train/plugins/profile/2022_12_09_12_50_45/DESKTOP-DHTLKO8.kernel_stats.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
+size 0

logs/cnn/train/plugins/profile/2022_12_09_12_50_45/DESKTOP-DHTLKO8.overview_page.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7aaa59e1543d35cc6a8f50d3f580a879a87180babc1b6e2b69b978bb2ab6f38d
+size 3011

logs/cnn/train/plugins/profile/2022_12_09_12_50_45/DESKTOP-DHTLKO8.tensorflow_stats.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f166e753702a5d9da8a493bcf02ff395c5a7ae1a9bf5589dbfaa0cef11481527
+size 36508

logs/cnn/train/plugins/profile/2022_12_09_12_50_45/DESKTOP-DHTLKO8.trace.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9952c168c5fb9d1b751c9bae0fc57d43a459de0fa3dfebaa5de609b226b3b12c
+size 4825

logs/cnn/validation/events.out.tfevents.1670589116.DESKTOP-DHTLKO8.21804.1555.v2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6d2b1df09e3974d460735f0500e967a69c4818f9b0bb0213ae30f671aa51c59
+size 13652

logs/cnn/validation/events.out.tfevents.1670589916.DESKTOP-DHTLKO8.4388.1603.v2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80abcf0e567063e5fc9db3ebba2ebb8eac425db9d9002e1537a73b05d0d4893b
+size 4336

logs/cnn/validation/events.out.tfevents.1670590254.DESKTOP-DHTLKO8.29980.1655.v2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:373f9e8828bbd0704fcdd803158b3adcf00a13618c4d02dcf6d27a8a4875f96c
+size 50780

main.py

@@ -0,0 +1,142 @@
+# Press Mayús+F10 to execute it or replace it with your code.
+# Press Double Shift to search everywhere for classes, files, tool windows, actions, and settings.
+
+import os
+import cv2
+import random
+import numpy as np
+import tensorflow as tf
+import matplotlib.pyplot as plt
+from tensorflow.keras.callbacks import TensorBoard
+from tensorflow.keras.preprocessing.image import ImageDataGenerator
+
+training_images = './Dataset/training'
+validation_images = './Dataset/validation'
+
+training_images_list = os.listdir(training_images)
+validation_images_list = os.listdir(validation_images)
+
+IMAGE_SIZE = 100
+
+# Get images from dataset
+def get_dataset_image(is_training_data):
+    images = []
+    tags = []
+    data = []
+    count = 0
+    image_list = validation_images_list
+    image_rute = validation_images
+
+    if is_training_data:
+        image_list = training_images_list
+        image_rute = training_images
+
+    for dir_name in image_list:
+        name = image_rute + '/' + dir_name
+
+        for file_name in os.listdir(name):
+            tags.append(count)
+            img = cv2.imread(name + '/' + file_name, 0)
+
+            if img is None:
+                print('Wrong path:', name + '/' + file_name)
+            else:
+                img = cv2.resize(img, (IMAGE_SIZE, IMAGE_SIZE), interpolation=cv2.INTER_CUBIC)
+                img = img.reshape(IMAGE_SIZE, IMAGE_SIZE, 1)
+                data.append([img, count])
+                images.append(img)
+
+        count = count + 1
+
+    return images, tags, data, count
+
+
+# Normalize images - white and black
+def normalize_images(images):
+    new_images = np.array(images).astype(float) / 255
+    return new_images
+
+
+# Avoid over fitting
+def avoid_over_fitting(images, tags):
+    rotation_range = random.randint(0, 90)
+    width_shift_range = random.uniform(0, 1)
+    height_shift_range = random.uniform(0, 1)
+    shear_range = random.randint(0, 25)
+
+    img_train_gen = ImageDataGenerator(
+        rotation_range=rotation_range,
+        width_shift_range=width_shift_range,
+        height_shift_range=height_shift_range,
+        shear_range=shear_range,
+        zoom_range=[0.5, 1.5],
+        vertical_flip=True,
+        horizontal_flip=True
+    )
+
+    img_train_gen.fit(images)
+    img_train = img_train_gen.flow(images, tags, batch_size=38)
+
+    return img_train
+
+
+# Training lists
+train_images, train_tags, train_data, train_count = get_dataset_image(True)
+
+# Validation lists
+val_images, val_tags, val_data, val_count = get_dataset_image(False)
+
+print('Read images finalized!')
+
+# Normalize
+train_images = normalize_images(train_images)
+val_images = normalize_images(val_images)
+
+train_tags = np.array(train_tags)
+val_tags = np.array(val_tags)
+
+img_to_train = avoid_over_fitting(train_images, train_tags)
+
+# Set layers and config CNN
+CNN_model = tf.keras.models.Sequential([
+    tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(IMAGE_SIZE, IMAGE_SIZE, 1)),
+    tf.keras.layers.MaxPooling2D(2, 2),
+    tf.keras.layers.Conv2D(64, (3, 3), activation='relu'),
+    tf.keras.layers.MaxPooling2D(2, 2),
+    tf.keras.layers.Conv2D(64, (3, 3), activation='relu'),
+    tf.keras.layers.MaxPooling2D(2, 2),
+    tf.keras.layers.Conv2D(128, (3, 3), activation='relu'),
+    tf.keras.layers.MaxPooling2D(2, 2),
+
+    # Clasification dense layers
+    tf.keras.layers.Dropout(0.2),
+    tf.keras.layers.Flatten(),
+    tf.keras.layers.Dense(250, activation='relu'),
+    tf.keras.layers.Dense(1, activation='sigmoid')
+])
+
+CNN_model.compile(
+    optimizer='adam',
+    loss='binary_crossentropy',
+    metrics=['accuracy']
+)
+
+
+# Train model
+BoardCNN = TensorBoard(log_dir='./logs/cnn')
+CNN_model.fit(
+    img_to_train,
+    batch_size=38,
+    validation_data=(val_images, val_tags),
+    epochs=500,
+    callbacks=[BoardCNN],
+    steps_per_epoch=int(np.ceil(len(train_images)/float(38))),
+    validation_steps=int(np.ceil(len(val_images)/float(38)))
+)
+
+
+# Save model
+CNN_model.save('./saves/dogs-cats.h5')
+CNN_model.save_weights('./saves/wights-dogs-cats.h5')
+
+print("Finish!")

predict.py

@@ -0,0 +1,51 @@
+import tensorflow as tf
+import cv2
+import numpy as np
+from keras_preprocessing.image import img_to_array
+
+# Get model
+model = './saves/dogs-cats.h5'
+
+IMAGE_SIZE = 100
+
+CNN_MODEL = tf.keras.models.load_model(model)
+weight_model = CNN_MODEL.get_weights()
+CNN_MODEL.set_weights(weight_model)
+
+# CAM
+cap = cv2.VideoCapture(0)
+
+while True:
+    ret, frame = cap.read()
+
+    # Parse to gray
+    img = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+    img = cv2.resize(img, (IMAGE_SIZE, IMAGE_SIZE), interpolation=cv2.INTER_CUBIC)
+
+    # Img normalized
+    img = np.array(img).astype(float) / 255
+
+    # Parse to 2D array
+    image = img_to_array(img)
+    image = np.expand_dims(image, axis=0)
+
+    # Predict
+    prediction = CNN_MODEL.predict(image)
+    prediction = prediction[0][0]
+
+    print(prediction)
+
+    # Classification
+    if prediction <= 0.5:
+        cv2.putText(frame, "Gato", (200, 70), cv2.FONT_HERSHEY_SIMPLEX, 3, (236, 19, 180), 6)
+    else:
+        cv2.putText(frame, "Perro", (200, 70), cv2.FONT_HERSHEY_SIMPLEX, 3, (20, 106, 231), 6)
+
+    cv2.imshow("CNN", frame)
+
+    t = cv2.waitKey(1)
+    if t == 25:
+        break
+
+cv2.destroyAllWindows()
+cap.release()

saves/dogs-cats.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:824d2473e0ff51bc5009ea63dac3cd2d4b420124576cda3641c81b6824aeae63
+size 7769168

saves/wights-dogs-cats.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2e7411658c25790f3bba05ee2e8e06619001d66fc01fe7f719291c73e352eea
+size 2597952