File size: 1,479 Bytes
ebc32f0
 
 
 
 
 
3903f4f
 
ebc32f0
 
99411de
7e9b170
d1b96ae
ebc32f0
 
 
 
 
 
 
 
 
 
 
 
 
3903f4f
 
 
 
 
ebc32f0
 
 
 
3903f4f
 
ebc32f0
 
 
3903f4f
ebc32f0
3903f4f
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
from functools import lru_cache

from huggingface_hub import hf_hub_download
from imgutils.data import ImageTyping, load_image, rgb_encode

from onnx_ import _open_onnx_model
from plot import detection_visualize
from yolo_ import _image_preprocess, _data_postprocess

_PERSON_MODELS = [
    'person_detect_plus_best_m.onnx',
    'person_detect_best_m.onnx',
    'person_detect_best_x.onnx',
    'person_detect_best_s.onnx',
]
_DEFAULT_PERSON_MODEL = _PERSON_MODELS[0]


@lru_cache()
def _open_person_detect_model(model_name):
    return _open_onnx_model(hf_hub_download(
        'deepghs/imgutils-models',
        f'person_detect/{model_name}'
    ))


_LABELS = ['person']


def detect_person(image: ImageTyping, level: str = 's', max_infer_size=1216,
                  conf_threshold: float = 0.3, iou_threshold: float = 0.5):
    image = load_image(image, mode='RGB')
    new_image, old_size, new_size = _image_preprocess(image, max_infer_size)

    data = rgb_encode(new_image)[None, ...]
    output, = _open_person_detect_model(level).run(['output0'], {'images': data})
    return _data_postprocess(output[0], conf_threshold, iou_threshold, old_size, new_size, _LABELS)


def _gr_detect_person(image: ImageTyping, model_name: str, max_infer_size=1216,
                      conf_threshold: float = 0.3, iou_threshold: float = 0.5):
    ret = detect_person(image, model_name, max_infer_size, conf_threshold, iou_threshold)
    return detection_visualize(image, ret, _LABELS)