Untitled-1.py

# %%
import webbrowser
import openai
import os
import cv2 as cv
import numpy as np
import mediapipe as mp
import matplotlib.pyplot as plt
import math
import imutils

# %%


def execute():
    def rotateImage(face):
        face_data = face.location_data
        rx = face_data.relative_keypoints[0].x
        ry = face_data.relative_keypoints[0].y
        lx = face_data.relative_keypoints[1].x
        ly = face_data.relative_keypoints[1].y
        deltaY = abs(ly - ry)
        deltaX = abs(lx - rx)
        rotationAngle = np.rad2deg(math.tan(deltaY/deltaX))
        rotatedImage = imutils.rotate(
            sample_img[:, :, ::-1].copy(), rotationAngle)
        return rotatedImage

    mp_face_detection = mp.solutions.face_detection
    face_detection = mp_face_detection.FaceDetection(
        model_selection=1, min_detection_confidence=0.5)
    mp_drawing = mp.solutions.drawing_utils
    sample_img = cv.imread('untitled.jpeg')

    plt.figure(figsize=[10, 10])

    face_detection_results = face_detection.process(sample_img[:, :, ::-1])
    if not (hasattr(face_detection_results, 'detections') and len(face_detection_results.detections) > 0):
        return

    rotatedImage = rotateImage(face_detection_results.detections[0])

    rotated_face_detection_results = face_detection.process(
        rotatedImage[:, :, ::-1])

    print('dddd')
    boundingBox = rotated_face_detection_results.detections[0].location_data.relative_bounding_box
    xmin = int(boundingBox.xmin * sample_img.shape[1])
    ymin = int(boundingBox.ymin * sample_img.shape[0])
    width = int(boundingBox.width * sample_img.shape[1])
    height = int(boundingBox.height * sample_img.shape[0])

    rotatedImage = rotatedImage[ymin:ymin+height, xmin:xmin+height]
    print(f'{xmin},{ymin},{width},{height}')

    background = np.zeros((1024, 1024, 4), np.uint8)

    x_offset = (background.shape[0] - rotatedImage.shape[0]) // 2
    y_offset = (background.shape[1] - rotatedImage.shape[1]) // 2

    alpha = np.sum(rotatedImage, axis=-1) > 0
    alpha = np.uint8(alpha * 255)
    rotatedRGBAImage = np.dstack((rotatedImage, alpha))

    background[y_offset:y_offset+rotatedRGBAImage.shape[0],
               x_offset:x_offset+rotatedRGBAImage.shape[1]] = rotatedRGBAImage

    plt.title("Resultant Image")
    plt.axis('off')
    plt.imshow(background)
    plt.show()
    cv.imwrite('image.png', cv.cvtColor(background, cv.COLOR_RGBA2BGRA))


execute()

# %% [markdown]
# # DALLE Portion
def generateHeadshot(amt):
# %%
    print("hello")
    openai.api_key = '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'

# %%
    response = openai.Image.create_edit(
        image=open("image.png", "rb"),
        prompt="A Professional headshot of one person you should hire",
        n=amt,
        size="1024x1024"
    )

    for r in response['data']:
        print(r)

import webbrowser 
new = 2 # open in a new tab, if possible

# open a public URL, in this case, the webbrowser docs
url = "http://docs.python.org/library/webbrowser.html"
webbrowser.get(using='google-chrome').open(url,new=new)