Incorrect max_val Parameter in peak_signal_noise_ratio function for Normalized Images

[sidd1915]

The peak_signal_noise_ratio function in the current implementation uses max_val=255.0. However, the images are normalized to the range [0, 1] in the read_image function:
image = tf.cast(image, dtype=tf.float32) / 255.0
Since the pixel values of y_true and y_pred are in the range [0, 1], the max_val parameter in tf.image.psnr should be set to 1.0 instead of 255.0 to correctly compute the PSNR.