Spaces:
Build error
Build error
File size: 2,089 Bytes
5b87ea7 b05691d 5b87ea7 b05691d 5b87ea7 ea56f13 5b87ea7 eeb64db 5b87ea7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 |
import cv2
import numpy as np
import urllib
import tensorflow as tf
from tensorflow.keras import layers, models, optimizers
import gradio as gr
import os
# Load pre-trained model
model = tf.keras.models.load_model('facial_condition_model.h5')
# Function to preprocess image for model input
def preprocess_image(image):
# Convert image to grayscale
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
# Resize image to 128x128 pixels
resized = cv2.resize(gray, (128, 128), interpolation=cv2.INTER_AREA)
# Convert image to float32 and normalize pixel values
normalized = np.float32(resized) / 255.0
# Add batch dimension to image
batched = np.expand_dims(normalized, axis=0)
return batched
# Function to calculate facial condition scores
def calculate_scores(image):
# Preprocess image for model input
preprocessed_image = preprocess_image(image)
# Use pre-trained model to predict facial condition scores
score = model.predict(preprocessed_image)[0][0]
# Calculate wrinkle score
wrinkle_score = 1 - (1 / (1 + np.exp(-5*(score-50)/50)))
# Calculate laxity score
laxity_score = np.tanh(score / 10)
# Calculate facial age
facial_age = int(np.round(score))
# Calculate final score
final_score = np.mean([wrinkle_score, laxity_score, facial_age])
return wrinkle_score, laxity_score, facial_age, final_score
# Define Gradio interface
inputs = gr.inputs.Image(label="Input Image")
outputs = [gr.outputs.Label(label="Wrinkle Score"),
gr.outputs.Label(label="Laxity Score"),
gr.outputs.Label(label="Facial Age"),
gr.outputs.Label(label="Final Score")]
title = "Facial Condition Evaluator"
description = "Upload a photo to evaluate your facial condition. This tool will calculate your wrinkle score, laxity score, facial age, and a final score based on a deep learning model."
examples = [["before.jpg"],
["after.jpg"]]
iface = gr.Interface(calculate_scores, inputs, outputs, title=title, description=description, examples=examples)
# Launch Gradio
iface.launch() |