135 lines
4.6 KiB
Python
135 lines
4.6 KiB
Python
|
|
|
|
import numpy as np
|
|
import os
|
|
import sys
|
|
import os.path as osp
|
|
|
|
sys.path.append(R"../0_strongReid")
|
|
import torchvision.transforms as T
|
|
import onnxruntime
|
|
import argparse
|
|
import torch
|
|
import torch.onnx
|
|
from PIL import Image
|
|
import PIL.Image as pil_image
|
|
from PIL import ImageDraw
|
|
from config import cfg
|
|
from modeling import build_model
|
|
from modeling.baseline import Baseline
|
|
normalize_transform = T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
|
|
|
|
IMG_PATH = "../3_deploy/modelzoo/strongReid/io/input"
|
|
TRACED_PATH = "../2_compile/fmodel/strongReid_256x128.onnx"
|
|
|
|
def read_image(img_path):
|
|
"""Keep reading image until succeed.
|
|
This can avoid IOError incurred by heavy IO process."""
|
|
got_img = False
|
|
if not osp.exists(img_path):
|
|
raise IOError("{} does not exist".format(img_path))
|
|
while not got_img:
|
|
try:
|
|
img = Image.open(img_path).convert('RGB')
|
|
got_img = True
|
|
except IOError:
|
|
print("IOError incurred when reading '{}'. Will redo. Don't worry. Just chill.".format(img_path))
|
|
pass
|
|
return img
|
|
|
|
|
|
transform = T.Compose([
|
|
T.Resize([256, 128]),
|
|
T.ToTensor(),
|
|
normalize_transform
|
|
])
|
|
def to_numpy(tensor):
|
|
return tensor.detach().cpu().numpy() if tensor.requires_grad else tensor.cpu().numpy()
|
|
|
|
def main():
|
|
|
|
from utils.data_infer import make_val_data_loader
|
|
parser = argparse.ArgumentParser(description="ReID Baseline Inference")
|
|
parser.add_argument(
|
|
"--config_file", default="iconfigs/softmax_triplet_with_center_self.yml", help="path to config file", type=str
|
|
)
|
|
parser.add_argument("opts", help="Modify config options using the command-line", default=None,
|
|
nargs=argparse.REMAINDER)
|
|
args = parser.parse_args()
|
|
cfg.merge_from_file(args.config_file)
|
|
cfg.merge_from_list(args.opts)
|
|
cfg['DATASETS']['ROOT_DIR'] = IMG_PATH
|
|
|
|
cfg.freeze()
|
|
|
|
prenum_classes = 751
|
|
val_loader, num_query, num_classes = make_val_data_loader(cfg)
|
|
# model = build_model(cfg, prenum_classes)
|
|
# checkpoint = torch.load(cfg.TEST.WEIGHT, map_location='cpu')
|
|
# model.load_state_dict(checkpoint)
|
|
# model.eval()
|
|
model = onnxruntime.InferenceSession(TRACED_PATH,providers=['CPUExecutionProvider'])
|
|
|
|
alldata = []
|
|
allfeats = []
|
|
allpids = []
|
|
allcamids = []
|
|
TRACED_ON = True
|
|
for batch in val_loader:
|
|
data, pids, camids = batch
|
|
im = transform(data[0]).unsqueeze(0)
|
|
inputs = {model.get_inputs()[0].name: to_numpy(im)}
|
|
feat = model.run(None, inputs)
|
|
feat = torch.Tensor(np.array(feat))[0]
|
|
# if TRACED_ON:
|
|
# torch.onnx.export(model, im, TRACED_PATH, opset_version=11, verbose=True)
|
|
# TRACED_ON = False
|
|
alldata.append(data[0])
|
|
allfeats.append(feat)
|
|
allpids.extend(np.asarray(pids))
|
|
allcamids.extend(np.asarray(camids))
|
|
|
|
feats = torch.cat(allfeats, dim=0)
|
|
if cfg.TEST.FEAT_NORM == 'yes':
|
|
print("The test feature is normalized")
|
|
feats = torch.nn.functional.normalize(feats, dim=1, p=2)
|
|
# query
|
|
qf = feats[:num_query]
|
|
# gallery
|
|
gf = feats[num_query:]
|
|
m, n = qf.shape[0], gf.shape[0]
|
|
distmat = torch.pow(qf, 2).sum(dim=1, keepdim=True).expand(m, n) + \
|
|
torch.pow(gf, 2).sum(dim=1, keepdim=True).expand(n, m).t()
|
|
distmat.addmm_(1, -2, qf, gf.t())
|
|
print("query2gallery_dist:",distmat.cpu().detach().numpy()[0])
|
|
|
|
top3_idx = torch.topk(distmat, k=3, largest=False).indices.cpu().numpy()
|
|
selected_images = []
|
|
|
|
for idx in top3_idx[0]:
|
|
selected_images.append(alldata[idx+ 1])
|
|
|
|
# 计算画布大小
|
|
image_width, image_height = selected_images[0].size # 假设所有图片大小相同
|
|
canvas_width = image_width * 3 # 3 张图片横向排列
|
|
canvas_height = image_height * 2 # 上半部分留空，下半部分放图片
|
|
|
|
# 创建空白画布
|
|
canvas = pil_image.new("RGB", (canvas_width, canvas_height), color=(255, 255, 255))
|
|
|
|
# 将图片粘贴到画布的下半部分
|
|
for i, img in enumerate(selected_images):
|
|
x = i * image_width # 横向位置
|
|
y = image_height # 纵向位置（从下半部分开始）
|
|
canvas.paste(img, (x, y))
|
|
|
|
canvas.paste(alldata[0], (0, 0))
|
|
draw = ImageDraw.Draw(canvas)
|
|
draw.text(xy=(0, 0), text='query', fill=(255, 0, 0))
|
|
draw.text(xy=(0, image_height), text='top3_match', fill=(255, 0, 0))
|
|
canvas.show()
|
|
canvas.save("output_2_save_infer.png")
|
|
|
|
if __name__ == '__main__':
|
|
main()
|