Can someone help me? i train my Machine Learning models with CIFAR10 datasets from tensorflow datasets, but i can’t increase my model accuracy above 80%….
can someone give me a suggestion?

import tensorflow as tf
import time
import tensorflow_datasets as tfds
from tensorflow.keras.datasets import cifar10
from tensorflow.keras.utils import to_categorical
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

(train_images, train_labels), (test_images, test_labels) = cifar10.load_data()

def normalize(train_images, test_images):
  normalized_train_dataset = tf.cast(train_images, tf.float32) / 255.0
  normalized_test_dataset = tf.cast(test_images, tf.float32) / 255.0
  return normalized_train_dataset, normalized_test_dataset

# Normalisasi Dataset
train_dataset, test_dataset = normalize(train_images, test_images)

def visualization(image, image_sample=2):
  for i in range (image_sample):

    print(f"bentuk gambar : {np.shape(image)}")
    print(f"bentuk data : {image[i].dtype}")
    print(f"Nilai Makzsimum : {np.max(image[i])}")
    print(f"Nilai Minimum : {np.min(image[i])}")


    plt.figure(figsize=(6,6))
    plt.imshow(image[i])
    plt.axis('off')
    plt.colorbar()
    plt.title("Gambar CIFAR-10")
    plt.grid(False)
    plt.show()

visualization(train_dataset)


train_labels = np.squeeze(train_labels)
test_labels = np.squeeze(test_labels)

print(f"Shape Of Train Label : {train_labels.shape}")
print(f"Shape Of Test_Label : {test_labels.shape}")

train_labels= to_categorical(train_labels, num_classes=10)
test_labels = to_categorical(test_labels, num_classes=10)

from tensorflow.keras.preprocessing.image import ImageDataGenerator

datagen = ImageDataGenerator(
    rotation_range=20,
    width_shift_range=0.2,
    height_shift_range=0.2,
    shear_range=0.2,
    zoom_range=0.2,
    horizontal_flip=True,
    fill_mode='nearest'
)

model = tf.keras.models.Sequential([
    tf.keras.layers.Conv2D(32, (3,3), padding='same', activation=tf.nn.relu, input_shape=(32, 32, 3)),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D((2,2), strides=2),

    tf.keras.layers.Conv2D(64, (3,3), padding='same', activation=tf.nn.relu),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D((2,2), strides=2),

    tf.keras.layers.Conv2D(128, (3,3), padding='same', activation=tf.nn.relu),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D((2,2), strides=2),

    tf.keras.layers.Conv2D(128, (3,3), padding='same', activation=tf.nn.relu),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D((2,2), strides=2),

    tf.keras.layers.Conv2D(512, (3,3), padding='same', activation=tf.nn.relu, kernel_regularizer=tf.keras.regularizers.l2(0.01)),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D((2,2), strides=2),

    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(512, activation=tf.nn.relu),
    tf.keras.layers.Dropout(0.3),

    tf.keras.layers.Dense(128, activation=tf.nn.relu),
    tf.keras.layers.Dropout(0.5),

    tf.keras.layers.Dense(10, activation=tf.nn.softmax)
])

model.summary()

model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),
              loss='categorical_crossentropy',
              metrics=['accuracy'])

early_stopping = tf.keras.callbacks.EarlyStopping(
    monitor='val_loss',
    patience=5,
    restore_best_weights=True
)

reducer_lr = tf.keras.callbacks.ReduceLROnPlateau(
    monitor='val_loss',
    factor=0.2,
    patience=3,
    verbose=1,
    min_lr=0.00001
)

callbacks = [early_stopping, reducer_lr]

start_time = time.time()

history = model.fit(datagen.flow(
    train_dataset,
    train_labels,
    batch_size=64),
    epochs=30,
    validation_data=(test_dataset, test_labels),
    callbacks=callbacks,
    verbose=1
    )

end_time = time.time()
training_time = end_time - start_time
print(f"Training time: {training_time/60:.2f} minutes")

model.save('hand_gesture_detect.keras')

# evaluasi model
loss_val, accuracy_val = model.evaluate(test_dataset, test_labels)
print(f"Loss : {loss_val}")
print(f"Accuracy Val : {accuracy_val}")

from tensorflow.keras.applications import ResNet50

base_model = ResNet50(weights='ImageNet', include_top=False, input_tensor=(32, 32, 3))

i have complexsify my model, but still the accuracy its just 77-80%, i have no idea how to increase my model accuracy