from megengine import hub
model = hub.load(
    "megengine/models",
    "deeplabv3plus_res101_voc_512size",
    pretrained=True,
)
model.eval()

所有预训练模型希望数据被正确预处理。模型要求输入BGR的图片, 建议缩放到512x512（Pascal VOC2012）或1024x2048（Cityscapes），最后做归一化处理 (均值为: [103.530, 116.280, 123.675], 标准差为: [57.375, 57.120, 58.395])。

下面是一段处理一张图片的样例代码。

# Download an example image from the megengine data website
import urllib
url, filename = ("https://data.megengine.org.cn/images/cat.jpg", "cat.jpg")
try: urllib.URLopener().retrieve(url, filename)
except: urllib.request.urlretrieve(url, filename)

# Read and pre-process the image
import cv2
import numpy as np
import megengine.data.transform as T
import megengine.functional as F

image = cv2.imread("cat.jpg")
orih, oriw = image.shape[:2]
transform = T.Compose([
    T.Resize((512, 512)),
    T.Normalize(mean=[103.530, 116.280, 123.675], std=[57.375, 57.120, 58.395]),  # BGR
    T.ToMode(),
])
processed_img = transform.apply(image)[np.newaxis]  # CHW -> 1CHW
pred = model(processed_img)
pred = pred.numpy().squeeze().argmax(axis=0)
pred = cv2.resize(pred.astype("uint8"), (oriw, orih), interpolation=cv2.INTER_LINEAR)

模型描述

目前我们提供了在Pascal VOC2012数据集或Cityscapes数据集上预训练的DeepLabV3+模型, 性能如下：

参考文献

• Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation Liang-Chieh Chen, Yukun Zhu, George Papandreou, Florian Schroff, and Hartwig Adam. European Conference on Computer Vision (ECCV), 2018.