import gradio as gr

from matplotlib import gridspec
import matplotlib.pyplot as plt
import numpy as np
from PIL import Image
import tensorflow as tf
from transformers import SegformerFeatureExtractor, TFSegformerForSemanticSegmentation

feature_extractor = SegformerFeatureExtractor.from_pretrained(
    "prem-timsina/segformer-b0-finetuned-food", from_pt=True
)
model = TFSegformerForSemanticSegmentation.from_pretrained(
    "prem-timsina/segformer-b0-finetuned-food", from_pt=True
)

def ade_palette():
    """ADE20K palette that maps each class to RGB values."""
    return [
        [93, 93, 93],
        [43, 240, 132],
        [139, 136, 240],
        [158, 83, 109],
        [6, 76, 151],
        [95, 170, 87],
        [273, 236, 139],
        [21, 155, 160],
        [188, 220, 166],
        [238, 96, 247],
        [223, 180, 221],
        [29, 97, 24],
        [3, 233, 248],
        [105, 118, 44],
        [203, 237, 63],
        [234, 100, 240],
        [19, 179, 164],
        [65, 22, 115],
        [111, 128, 194],
        [232, 41, 17],
        [11, 250, 159],
        [137, 163, 129],
        [212, 223, 210],
        [51, 37, 4],
        [37, 63, 239],
        [257, 180, 163],
        [172, 53, 105],
        [104, 150, 99],
        [80, 157, 133],
        [195, 104, 202],
        [42, 187, 110],
        [133, 225, 66],
        [132, 99, 213],
        [178, 248, 209],
        [93, 147, 60],
        [105, 109, 115],
        [26, 65, 115],
        [239, 52, 182],
        [242, 19, 204],
        [157, 101, 214],
        [248, 85, 198],
        [103, 198, 171],
        [44, 129, 75],
        [159, 32, 120],
        [155, 77, 71],
        [233, 231, 155],
        [135, 196, 206],
        [81, 53, 51],
        [134, 221, 213],
        [192, 27, 152],
        [127, 127, 194],
        [82, 161, 1],
        [71, 80, 161],
        [148, 9, 159],
        [91, 110, 124],
        [127, 157, 223],
        [25, 210, 232],
        [129, 0, 114],
        [231, 187, 138],
        [23, 17, 224],
        [25, 255, 29],
        [158, 19, 53],
        [157, 190, 176],
        [114, 140, 221],
        [46, 104, 87],
        [17, 114, 122],
        [221, 12, 229],
        [54, 20, 92],
        [215, 191, 252],
        [144, 127, 146],
        [141, 116, 77],
        [100, 89, 89],
        [104, 115, 249],
        [179, 212, 38],
        [140, 248, 179],
        [177, 230, 240],
        [219, 98, 8],
        [74, 219, 53],
        [161, 28, 243],
        [64, 57, 184],
        [147, 193, 113],
        [182, 15, 30],
        [151, 204, 109],
        [187, 76, 21],
        [118, 163, 155],
        [158, 30, 220],
        [227, 170, 63],
        [199, 186, 72],
        [0, 241, 168],
        [80, 150, 225],
        [237, 250, 4],
        [29, 210, 181],
        [176, 120, 81],
        [134, 47, 123],
        [240, 141, 130],
        [250, 41, 115],
        [29, 88, 143],
        [66, 151, 87],
        [241, 231, 144],
        [238, 107, 153],
        [181, 96, 220],
        [239, 122, 133],
        [205, 120, 21],
        [168, 12, 77],

    ]

labels_list = []

with open(r'labels.txt', 'r') as fp:
    for line in fp:
        labels_list.append(line[:-1])

colormap = np.asarray(ade_palette())

def label_to_color_image(label):
    if label.ndim != 2:
        raise ValueError("Expect 2-D input label")

    if np.max(label) >= len(colormap):
        raise ValueError("label value too large.")
    return colormap[label]

def draw_plot(pred_img, seg):
    fig = plt.figure(figsize=(20, 15))

    grid_spec = gridspec.GridSpec(1, 2, width_ratios=[6, 1])

    plt.subplot(grid_spec[0])
    plt.imshow(pred_img)
    plt.axis('off')
    LABEL_NAMES = np.asarray(labels_list)
    FULL_LABEL_MAP = np.arange(len(LABEL_NAMES)).reshape(len(LABEL_NAMES), 1)
    FULL_COLOR_MAP = label_to_color_image(FULL_LABEL_MAP)

    unique_labels = np.unique(seg.numpy().astype("uint8"))
    ax = plt.subplot(grid_spec[1])
    plt.imshow(FULL_COLOR_MAP[unique_labels].astype(np.uint8), interpolation="nearest")
    ax.yaxis.tick_right()
    plt.yticks(range(len(unique_labels)), LABEL_NAMES[unique_labels])
    plt.xticks([], [])
    ax.tick_params(width=0.0, labelsize=25)
    return fig

def sepia(input_img):
    input_img = Image.fromarray(input_img)

    inputs = feature_extractor(images=input_img, return_tensors="tf")
    outputs = model(**inputs)
    logits = outputs.logits

    logits = tf.transpose(logits, [0, 2, 3, 1])
    logits = tf.image.resize(
        logits, input_img.size[::-1]
    )  # We reverse the shape of `image` because `image.size` returns width and height.
    seg = tf.math.argmax(logits, axis=-1)[0]

    color_seg = np.zeros(
        (seg.shape[0], seg.shape[1], 3), dtype=np.uint8
    )  # height, width, 3
    for label, color in enumerate(colormap):
        color_seg[seg.numpy() == label, :] = color

    # Show image + mask
    pred_img = np.array(input_img) * 0.5 + color_seg * 0.5
    pred_img = pred_img.astype(np.uint8)

    fig = draw_plot(pred_img, seg)
    return fig

demo = gr.Interface(fn=sepia,
                    inputs=gr.Image(shape=(400, 600)),
                    outputs=['plot'],
                    examples=["food-1.jpg","food-2.jpg", "food-3.jpg", "food-4.jpg", "food-5.jpg", "food-6.jpg"],
                    allow_flagging='never')


demo.launch()