172 lines
6.1 KiB
Python
172 lines
6.1 KiB
Python
import os
|
|
from typing import Annotated
|
|
|
|
import cv2
|
|
import tqdm
|
|
import typer
|
|
import imageio
|
|
import numpy as np
|
|
from PIL import Image
|
|
|
|
import torch
|
|
import torch.nn.functional as F
|
|
|
|
from matanyone2.utils.download_util import load_file_from_url
|
|
from matanyone2.utils.inference_utils import gen_dilate, gen_erosion, read_frame_from_videos
|
|
from matanyone2.inference.inference_core import InferenceCore
|
|
from matanyone2.utils.get_default_model import get_matanyone2_model
|
|
from matanyone2.utils.device import get_default_device, safe_autocast_decorator
|
|
|
|
import warnings
|
|
warnings.filterwarnings("ignore")
|
|
|
|
app = typer.Typer(help="MatAnyone2 video matting inference")
|
|
|
|
|
|
@torch.inference_mode()
|
|
@safe_autocast_decorator()
|
|
def run_inference(input_path, mask_path, output_path, ckpt_path, n_warmup=10, r_erode=10,
|
|
r_dilate=10, suffix="", save_image=False, max_size=-1):
|
|
device = get_default_device()
|
|
|
|
# download ckpt for the first inference
|
|
pretrain_model_url = "https://github.com/pq-yang/MatAnyone2/releases/download/v1.0.0/matanyone2.pth"
|
|
ckpt_path = load_file_from_url(pretrain_model_url, 'pretrained_models')
|
|
|
|
# load MatAnyone model
|
|
matanyone2 = get_matanyone2_model(ckpt_path, device)
|
|
|
|
# init inference processor
|
|
processor = InferenceCore(matanyone2, cfg=matanyone2.cfg)
|
|
|
|
# inference parameters
|
|
r_erode = int(r_erode)
|
|
r_dilate = int(r_dilate)
|
|
n_warmup = int(n_warmup)
|
|
max_size = int(max_size)
|
|
|
|
# load input frames
|
|
vframes, fps, length, video_name = read_frame_from_videos(input_path)
|
|
repeated_frames = vframes[0].unsqueeze(0).repeat(n_warmup, 1, 1, 1)
|
|
vframes = torch.cat([repeated_frames, vframes], dim=0).float()
|
|
length += n_warmup
|
|
|
|
# resize if needed
|
|
if max_size > 0:
|
|
h, w = vframes.shape[-2:]
|
|
min_side = min(h, w)
|
|
if min_side > max_size:
|
|
new_h = int(h / min_side * max_size)
|
|
new_w = int(w / min_side * max_size)
|
|
vframes = F.interpolate(vframes, size=(new_h, new_w), mode="area")
|
|
print(f'Resize to {new_h}x{new_w} for processing...')
|
|
|
|
# set output paths
|
|
os.makedirs(output_path, exist_ok=True)
|
|
if suffix != "":
|
|
video_name = f'{video_name}_{suffix}'
|
|
if save_image:
|
|
os.makedirs(f'{output_path}/{video_name}', exist_ok=True)
|
|
os.makedirs(f'{output_path}/{video_name}/pha', exist_ok=True)
|
|
os.makedirs(f'{output_path}/{video_name}/fgr', exist_ok=True)
|
|
|
|
# load the first-frame mask
|
|
mask = Image.open(mask_path).convert('L')
|
|
mask = np.array(mask)
|
|
|
|
bgr = (np.array([120, 255, 155], dtype=np.float32) / 255).reshape((1, 1, 3))
|
|
objects = [1]
|
|
|
|
# [optional] erode & dilate
|
|
if r_dilate > 0:
|
|
mask = gen_dilate(mask, r_dilate, r_dilate)
|
|
if r_erode > 0:
|
|
mask = gen_erosion(mask, r_erode, r_erode)
|
|
|
|
mask = torch.from_numpy(mask).float().to(device)
|
|
|
|
if max_size > 0:
|
|
mask = F.interpolate(mask.unsqueeze(0).unsqueeze(0), size=(new_h, new_w), mode="nearest")
|
|
mask = mask[0, 0]
|
|
|
|
# inference start
|
|
phas = []
|
|
fgrs = []
|
|
for ti in tqdm.tqdm(range(length)):
|
|
image = vframes[ti]
|
|
image_np = np.array(image.permute(1, 2, 0))
|
|
image = (image / 255.).float().to(device)
|
|
|
|
if ti == 0:
|
|
output_prob = processor.step(image, mask, objects=objects)
|
|
output_prob = processor.step(image, first_frame_pred=True)
|
|
else:
|
|
if ti <= n_warmup:
|
|
output_prob = processor.step(image, first_frame_pred=True)
|
|
else:
|
|
output_prob = processor.step(image)
|
|
|
|
mask = processor.output_prob_to_mask(output_prob)
|
|
|
|
pha = mask.unsqueeze(2).cpu().numpy()
|
|
com_np = image_np / 255. * pha + bgr * (1 - pha)
|
|
|
|
if ti > (n_warmup - 1):
|
|
com_np = np.round(np.clip(com_np * 255.0, 0, 255)).astype(np.uint8)
|
|
pha = np.round(np.clip(pha * 255.0, 0, 255)).astype(np.uint8)
|
|
fgrs.append(com_np)
|
|
phas.append(pha)
|
|
if save_image:
|
|
cv2.imwrite(f'{output_path}/{video_name}/fgr/{str(ti - n_warmup).zfill(4)}.png',
|
|
com_np[..., [2, 1, 0]])
|
|
cv2.imwrite(f'{output_path}/{video_name}/pha/{str(ti - n_warmup).zfill(4)}.png',
|
|
pha)
|
|
|
|
phas = np.array(phas)
|
|
fgrs = np.array(fgrs)
|
|
|
|
imageio.mimwrite(f'{output_path}/{video_name}_fgr.mp4', fgrs, fps=fps, quality=7)
|
|
imageio.mimwrite(f'{output_path}/{video_name}_pha.mp4', phas, fps=fps, quality=7)
|
|
|
|
|
|
@app.command()
|
|
def main(
|
|
input_path: Annotated[str, typer.Option("-i", "--input-path",
|
|
help="Path of the input video or frame folder.")],
|
|
mask_path: Annotated[str, typer.Option("-m", "--mask-path",
|
|
help="Path of the first-frame segmentation mask.")],
|
|
output_path: Annotated[str, typer.Option("-o", "--output-path",
|
|
help="Output folder.")] = "results/",
|
|
ckpt_path: Annotated[str, typer.Option("-c", "--ckpt-path",
|
|
help="Path of the MatAnyone2 model.")] = "pretrained_models/matanyone2.pth",
|
|
warmup: Annotated[int, typer.Option("-w", "--warmup",
|
|
help="Number of warmup iterations for the first frame alpha prediction.")] = 10,
|
|
erode_kernel: Annotated[int, typer.Option("-e", "--erode-kernel",
|
|
help="Erosion kernel size on the input mask.")] = 10,
|
|
dilate_kernel: Annotated[int, typer.Option("-d", "--dilate-kernel",
|
|
help="Dilation kernel size on the input mask.")] = 10,
|
|
suffix: Annotated[str, typer.Option(
|
|
help="Suffix to specify different target when saving.")] = "",
|
|
save_image: Annotated[bool, typer.Option("--save-image",
|
|
help="Save output frames.")] = False,
|
|
max_size: Annotated[int, typer.Option(
|
|
help="Downsamples if min(w, h) exceeds this value. -1 means no limit.")] = -1,
|
|
):
|
|
"""Run MatAnyone2 video matting inference."""
|
|
run_inference(
|
|
input_path=input_path,
|
|
mask_path=mask_path,
|
|
output_path=output_path,
|
|
ckpt_path=ckpt_path,
|
|
n_warmup=warmup,
|
|
r_erode=erode_kernel,
|
|
r_dilate=dilate_kernel,
|
|
suffix=suffix,
|
|
save_image=save_image,
|
|
max_size=max_size,
|
|
)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
app()
|