File size: 3,557 Bytes
98a6605
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
import os
import sys
import numpy as np
import tensorflow as tf
import mediapy
from PIL import Image
import gradio as gr
from huggingface_hub import snapshot_download

# Clone the repository and add the path
os.system("git clone https://github.com/google-research/frame-interpolation")
sys.path.append("frame-interpolation")

# Import after appending the path
from eval import interpolator, util

def load_model(model_name):
    model = interpolator.Interpolator(snapshot_download(repo_id=model_name), None)
    return model

model_names = [
    "akhaliq/frame-interpolation-film-style",
    "NimaBoscarino/frame-interpolation_film_l1",
    "NimaBoscarino/frame_interpolation_film_vgg",
]

models = {model_name: load_model(model_name) for model_name in model_names}

ffmpeg_path = util.get_ffmpeg_path()
mediapy.set_ffmpeg(ffmpeg_path)

def resize(width, img):
    img = Image.fromarray(img)
    wpercent = (width / float(img.size[0]))
    hsize = int((float(img.size[1]) * float(wpercent)))
    img = img.resize((width, hsize), Image.LANCZOS)
    return img

def resize_and_crop(img_path, size, crop_origin="middle"):
    img = Image.open(img_path)
    img = img.resize(size, Image.LANCZOS)
    return img

def resize_img(img1, img2_path):
    img_target_size = Image.open(img1)
    img_to_resize = resize_and_crop(
        img2_path,
        (img_target_size.size[0], img_target_size.size[1]),  # set width and height to match img1
        crop_origin="middle"
    )
    img_to_resize.save('resized_img2.png')

def predict(frame1, frame2, frame3, frame4, frame5, frame6, times_to_interpolate, model_name):
    model = models[model_name]

    # Resize all frames
    frames = [resize(1080, frame) for frame in [frame1, frame2, frame3, frame4, frame5, frame6]]

    # Save and resize images
    for i, frame in enumerate(frames):
        frame.save(f"test{i+1}.png")
        if i > 0:  # Resize all except the first frame
            resize_img(f"test1.png", f"test{i+1}.png")

    input_frames = [f"test{i+1}.png" for i in range(6)]

    # Interpolate using the model
    interpolated_frames = list(util.interpolate_recursively_from_files(input_frames, times_to_interpolate, model))

    mediapy.write_video("out.mp4", interpolated_frames, fps=30)
    return "out.mp4"

title = "frame-interpolation"
description = "Gradio demo for FILM: Frame Interpolation for Large Scene Motion. To use it, simply upload your images and add the times to interpolate number or click on one of the examples to load them. Read more at the links below."
article = "<p style='text-align: center'><a href='https://film-net.github.io/' target='_blank'>FILM: Frame Interpolation for Large Motion</a> | <a href='https://github.com/google-research/frame-interpolation' target='_blank'>Github Repo</a></p>"
examples = [
    ['cat3.jpeg', 'cat4.jpeg', 'cat5.jpeg', 'cat6.jpeg', 'cat7.jpeg', 'cat8.jpeg', 2, model_names[0]],
    ['cat1.jpeg', 'cat2.jpeg', 'cat3.jpeg', 'cat4.jpeg', 'cat5.jpeg', 'cat6.jpeg', 2, model_names[1]],
]

gr.Interface(
    fn=predict,
    inputs=[
        gr.Image(label="First Frame"),
        gr.Image(label="Second Frame"),
        gr.Image(label="Third Frame"),
        gr.Image(label="Fourth Frame"),
        gr.Image(label="Fifth Frame"),
        gr.Image(label="Sixth Frame"),
        gr.Number(label="Times to Interpolate", value=2),
        gr.Dropdown(label="Model", choices=model_names),
    ],
    outputs=gr.Video(label="Interpolated Frames"),
    title=title,
    description=description,
    article=article,
    examples=examples,
).launch()