290 lines
12 KiB
Python
Raw Permalink Normal View History

2025-12-09 10:15:16 +08:00
import os
import cv2
import numpy as np
import onnxruntime as ort
def distance2box(points, distance, max_shape=None):
x1 = points[:, 0] - distance[:, 0]
y1 = points[:, 1] - distance[:, 1]
x2 = points[:, 0] + distance[:, 2]
y2 = points[:, 1] + distance[:, 3]
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1])
y1 = y1.clamp(min=0, max=max_shape[0])
x2 = x2.clamp(min=0, max=max_shape[1])
y2 = y2.clamp(min=0, max=max_shape[0])
return np.stack([x1, y1, x2, y2], axis=-1)
def distance2kps(points, distance, max_shape=None):
outputs = []
for i in range(0, distance.shape[1], 2):
p_x = points[:, i % 2] + distance[:, i]
p_y = points[:, i % 2 + 1] + distance[:, i + 1]
if max_shape is not None:
p_x = p_x.clamp(min=0, max=max_shape[1])
p_y = p_y.clamp(min=0, max=max_shape[0])
outputs.append(p_x)
outputs.append(p_y)
return np.stack(outputs, axis=-1)
class FaceDetector:
def __init__(self, onnx_path=None, session=None):
from onnxruntime import InferenceSession
self.session = session
self.batched = False
if self.session is None:
assert onnx_path is not None
assert os.path.exists(onnx_path)
self.session = InferenceSession(onnx_path,
providers=['CUDAExecutionProvider'])
self.nms_thresh = 0.4
self.center_cache = {}
input_cfg = self.session.get_inputs()[0]
input_shape = input_cfg.shape
if isinstance(input_shape[2], str):
self.input_size = None
else:
self.input_size = tuple(input_shape[2:4][::-1])
input_name = input_cfg.name
outputs = self.session.get_outputs()
if len(outputs[0].shape) == 3:
self.batched = True
output_names = []
for output in outputs:
output_names.append(output.name)
self.input_name = input_name
self.output_names = output_names
self.use_kps = False
self._num_anchors = 1
if len(outputs) == 6:
self.fmc = 3
self._feat_stride_fpn = [8, 16, 32]
self._num_anchors = 2
elif len(outputs) == 9:
self.fmc = 3
self._feat_stride_fpn = [8, 16, 32]
self._num_anchors = 2
self.use_kps = True
elif len(outputs) == 10:
self.fmc = 5
self._feat_stride_fpn = [8, 16, 32, 64, 128]
self._num_anchors = 1
elif len(outputs) == 15:
self.fmc = 5
self._feat_stride_fpn = [8, 16, 32, 64, 128]
self._num_anchors = 1
self.use_kps = True
def forward(self, x, score_thresh):
scores_list = []
bboxes_list = []
points_list = []
input_size = tuple(x.shape[0:2][::-1])
blob = cv2.dnn.blobFromImage(x,
1.0 / 128,
input_size,
(127.5, 127.5, 127.5), swapRB=True)
outputs = self.session.run(self.output_names, {self.input_name: blob})
input_height = blob.shape[2]
input_width = blob.shape[3]
fmc = self.fmc
for idx, stride in enumerate(self._feat_stride_fpn):
if self.batched:
scores = outputs[idx][0]
boxes = outputs[idx + fmc][0]
boxes = boxes * stride
else:
scores = outputs[idx]
boxes = outputs[idx + fmc]
boxes = boxes * stride
height = input_height // stride
width = input_width // stride
key = (height, width, stride)
if key in self.center_cache:
anchor_centers = self.center_cache[key]
else:
anchor_centers = np.stack(np.mgrid[:height, :width][::-1], axis=-1)
anchor_centers = anchor_centers.astype(np.float32)
anchor_centers = (anchor_centers * stride).reshape((-1, 2))
if self._num_anchors > 1:
anchor_centers = np.stack([anchor_centers] * self._num_anchors, axis=1)
anchor_centers = anchor_centers.reshape((-1, 2))
if len(self.center_cache) < 100:
self.center_cache[key] = anchor_centers
pos_indices = np.where(scores >= score_thresh)[0]
bboxes = distance2box(anchor_centers, boxes)
pos_scores = scores[pos_indices]
pos_bboxes = bboxes[pos_indices]
scores_list.append(pos_scores)
bboxes_list.append(pos_bboxes)
return scores_list, bboxes_list
def detect(self, image, input_size=None, score_threshold=0.5, max_num=0, metric='default'):
assert input_size is not None or self.input_size is not None
input_size = self.input_size if input_size is None else input_size
image_ratio = float(image.shape[0]) / image.shape[1]
model_ratio = float(input_size[1]) / input_size[0]
if image_ratio > model_ratio:
new_height = input_size[1]
new_width = int(new_height / image_ratio)
else:
new_width = input_size[0]
new_height = int(new_width * image_ratio)
det_scale = float(new_height) / image.shape[0]
resized_img = cv2.resize(image, (new_width, new_height))
det_img = np.zeros((input_size[1], input_size[0], 3), dtype=np.uint8)
det_img[:new_height, :new_width, :] = resized_img
scores_list, bboxes_list = self.forward(det_img, score_threshold)
scores = np.vstack(scores_list)
scores_ravel = scores.ravel()
order = scores_ravel.argsort()[::-1]
bboxes = np.vstack(bboxes_list) / det_scale
pre_det = np.hstack((bboxes, scores)).astype(np.float32, copy=False)
pre_det = pre_det[order, :]
keep = self.nms(pre_det)
det = pre_det[keep, :]
if 0 < max_num < det.shape[0]:
area = (det[:, 2] - det[:, 0]) * (det[:, 3] - det[:, 1])
img_center = image.shape[0] // 2, image.shape[1] // 2
offsets = np.vstack([(det[:, 0] + det[:, 2]) / 2 - img_center[1],
(det[:, 1] + det[:, 3]) / 2 - img_center[0]])
offset_dist_squared = np.sum(np.power(offsets, 2.0), 0)
if metric == 'max':
values = area
else:
values = area - offset_dist_squared * 2.0 # some extra weight on the centering
index = np.argsort(values)[::-1] # some extra weight on the centering
index = index[0:max_num]
det = det[index, :]
return det
def nms(self, outputs):
thresh = self.nms_thresh
x1 = outputs[:, 0]
y1 = outputs[:, 1]
x2 = outputs[:, 2]
y2 = outputs[:, 3]
scores = outputs[:, 4]
order = scores.argsort()[::-1]
areas = (x2 - x1 + 1) * (y2 - y1 + 1)
keep = []
while order.size > 0:
i = order[0]
keep.append(i)
xx1 = np.maximum(x1[i], x1[order[1:]])
yy1 = np.maximum(y1[i], y1[order[1:]])
xx2 = np.minimum(x2[i], x2[order[1:]])
yy2 = np.minimum(y2[i], y2[order[1:]])
w = np.maximum(0.0, xx2 - xx1 + 1)
h = np.maximum(0.0, yy2 - yy1 + 1)
inter = w * h
ovr = inter / (areas[i] + areas[order[1:]] - inter)
indices = np.where(ovr <= thresh)[0]
order = order[indices + 1]
return keep
class HSEmotionRecognizer:
def __init__(self, path):
self.img_size = 260
self.mean = [0.485, 0.456, 0.406]
self.std = [0.229, 0.224, 0.225]
2025-12-09 11:28:23 +08:00
# self.idx_to_class = {0: 'ANGER',
# 1: 'DISGUST',
# 2: 'FEAR',
# 3: 'HAPPINESS',
# 4: 'NEUTRAL',
# 5: 'SADNESS',
# 6: 'SURPRISE'}
self.idx_to_class = {
0: '愤怒',
1: '厌恶',
2: '恐惧',
3: '快乐',
4: '中性',
5: '悲伤',
6: '惊讶'
}
2025-12-09 10:15:16 +08:00
self.ort_session = ort.InferenceSession(path, providers=['CUDAExecutionProvider'])
def preprocess(self, img):
x = cv2.resize(img, (self.img_size, self.img_size)) / 255
for i in range(3):
x[..., i] = (x[..., i] - self.mean[i]) / self.std[i]
return x.transpose(2, 0, 1).astype("float32")[np.newaxis, ...]
def predict_emotions(self, face_img, logits=True):
scores = self.ort_session.run(None, {"input": self.preprocess(face_img)})[0][0]
x = scores
pred = np.argmax(x)
if not logits:
e_x = np.exp(x - np.max(x)[np.newaxis])
e_x = e_x / e_x.sum()[None]
scores = e_x
return self.idx_to_class[pred], scores,pred
def predict_multi_emotions(self, face_img_list, logits=True):
images = np.concatenate([self.preprocess(face_img) for face_img in face_img_list], axis=0)
scores = self.ort_session.run(None, {"input": images})[0]
pred = np.argmax(scores, axis=1)
x = scores
if not logits:
e_x = np.exp(x - np.max(x, axis=1)[:, np.newaxis])
e_x = e_x / e_x.sum(axis=1)[:, None]
scores = e_x
return [self.idx_to_class[pred] for pred in pred], scores
def draw_rounded_bar(image, top_left, bottom_right, color, radius):
top_left = (int(top_left[0]), int(top_left[1]))
bottom_right = (int(bottom_right[0]), int(bottom_right[1]))
image_height, image_width = image.shape[:2]
bottom_right = (min(bottom_right[0], image_width), min(bottom_right[1], image_height))
cv2.rectangle(image, (top_left[0] + radius, top_left[1]), (bottom_right[0] - radius, bottom_right[1]), color,
thickness=2)
cv2.rectangle(image, (top_left[0], top_left[1] + radius), (bottom_right[0], bottom_right[1] - radius), color,
thickness=2)
cv2.circle(image, (top_left[0] + radius, top_left[1] + radius), radius, color, -1)
cv2.circle(image, (bottom_right[0] - radius, top_left[1] + radius), radius, color, -1)
cv2.circle(image, (top_left[0] + radius, bottom_right[1] - radius), radius, color, -1)
cv2.circle(image, (bottom_right[0] - radius, bottom_right[1] - radius), radius, color, -1)
def draw_emotion_bars(model, frame, scores, top_left, bar_height=25, width=150, spacing=10):
emotions = list(model.idx_to_class.values())
start_x, start_y = top_left
overlay = frame.copy()
overlay_height = len(emotions) * (bar_height + spacing) - spacing
cv2.rectangle(overlay, (start_x, start_y), (start_x + width, start_y + overlay_height), (0, 0, 0, 128), -1)
cv2.addWeighted(overlay, 0.4, frame, 1 - 0.4, 0, frame)
for i, (emotion, score) in enumerate(zip(emotions, scores)):
bar_length = int(score * width)
bar_top_left = (start_x, start_y + i * (bar_height + spacing))
bar_bottom_right = (start_x + bar_length, start_y + (i * (bar_height + spacing)) + bar_height)
draw_rounded_bar(frame, bar_top_left, bar_bottom_right, (0, 255, 0), radius=int(bar_height / 3))
text_position = (start_x + width + 5, bar_top_left[1] + bar_height // 2 + spacing // 4)
cv2.putText(frame, emotion, text_position, cv2.FONT_HERSHEY_COMPLEX, 0.4, (66, 15, 7), 1)