face_detector.py

import cv2
import os
import argparse
import numpy as np
import timeit
from tqdm import tqdm
import pickle
import random
from face import Face
import onnxruntime
onnxruntime.set_default_logger_severity(3)


def distance2bbox(points, distance, max_shape=None):
    """Decode distance prediction to bounding box.

    Args:
        points (Tensor): Shape (n, 2), [x, y].
        distance (Tensor): Distance from the given point to 4
            boundaries (left, top, right, bottom).
        max_shape (tuple): Shape of the image.

    Returns:
        Tensor: Decoded bboxes.
    """
    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):
    """Decode distance prediction to bounding box.

    Args:
        points (Tensor): Shape (n, 2), [x, y].
        distance (Tensor): Distance from the given point to 4
            boundaries (left, top, right, bottom).
        max_shape (tuple): Shape of the image.

    Returns:
        Tensor: Decoded bboxes.
    """
    preds = []
    for i in range(0, distance.shape[1], 2):
        px = points[:, i%2] + distance[:, i]
        py = points[:, i%2+1] + distance[:, i+1]
        if max_shape is not None:
            px = px.clamp(min=0, max=max_shape[1])
            py = py.clamp(min=0, max=max_shape[0])
        preds.append(px)
        preds.append(py)
    return np.stack(preds, axis=-1)


class SCRFD:
    def __init__(self, model_file=None, session=None):
        import onnxruntime
        self.model_file = model_file
        self.session = session
        self.taskname = 'detection'
        self.batched = False
        if self.session is None:
            assert self.model_file is not None
            assert os.path.exists(self.model_file)
            self.session = onnxruntime.InferenceSession(self.model_file, None)
        self.center_cache = {}
        self.nms_thresh = 0.4
        self.det_thresh = 0.5
        self._init_vars()

    def _init_vars(self):
        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
        self.input_shape = input_shape
        outputs = self.session.get_outputs()
        if len(outputs[0].shape) == 3:
            self.batched = True
        output_names = []
        for o in outputs:
            output_names.append(o.name)
        self.input_name = input_name
        self.output_names = output_names
        self.input_mean = 127.5
        self.input_std = 128.0
        self.use_kps = False
        self._anchor_ratio = 1.0
        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 prepare(self, ctx_id, **kwargs):
        if ctx_id<0:
            self.session.set_providers(['CPUExecutionProvider'])
        print(f'Applied providers: {self.session._providers}, with options: {self.session._provider_options}')
        nms_thresh = kwargs.get('nms_thresh', None)
        if nms_thresh is not None:
            self.nms_thresh = nms_thresh
        det_thresh = kwargs.get('det_thresh', None)
        if det_thresh is not None:
            self.det_thresh = det_thresh
        input_size = kwargs.get('input_size', None)
        if input_size is not None:
            if self.input_size is not None:
                print('warning: det_size is already set in scrfd model, ignore')
            else:
                self.input_size = input_size

    def forward(self, img, threshold):
        scores_list = []
        bboxes_list = []
        kpss_list = []
        input_size = tuple(img.shape[0:2][::-1])
        blob = cv2.dnn.blobFromImage(img, 1.0/self.input_std, input_size, (self.input_mean, self.input_mean, self.input_mean), swapRB=True)
        net_outs = 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 model support batch dim, take first output
            if self.batched:
                scores = net_outs[idx][0]
                bbox_preds = net_outs[idx + fmc][0]
                bbox_preds = bbox_preds * stride
                if self.use_kps:
                    kps_preds = net_outs[idx + fmc * 2][0] * stride
            # If model doesn't support batching take output as is
            else:
                scores = net_outs[idx]
                bbox_preds = net_outs[idx + fmc]
                bbox_preds = bbox_preds * stride
                if self.use_kps:
                    kps_preds = net_outs[idx + fmc * 2] * stride

            height = input_height // stride
            width = input_width // stride
            K = height * width
            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).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).reshape( (-1,2) )
                if len(self.center_cache)<100:
                    self.center_cache[key] = anchor_centers

            pos_inds = np.where(scores>=threshold)[0]
            bboxes = distance2bbox(anchor_centers, bbox_preds)
            pos_scores = scores[pos_inds]
            pos_bboxes = bboxes[pos_inds]
            scores_list.append(pos_scores)
            bboxes_list.append(pos_bboxes)
            if self.use_kps:
                kpss = distance2kps(anchor_centers, kps_preds)
                kpss = kpss.reshape( (kpss.shape[0], -1, 2) )
                pos_kpss = kpss[pos_inds]
                kpss_list.append(pos_kpss)
        return scores_list, bboxes_list, kpss_list

    def detect(self, img, input_size = None, 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
            
        im_ratio = float(img.shape[0]) / img.shape[1]
        model_ratio = float(input_size[1]) / input_size[0]
        if im_ratio>model_ratio:
            new_height = input_size[1]
            new_width = int(new_height / im_ratio)
        else:
            new_width = input_size[0]
            new_height = int(new_width * im_ratio)
        det_scale = float(new_height) / img.shape[0]
        resized_img = cv2.resize(img, (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, kpss_list = self.forward(det_img, self.det_thresh)

        scores = np.vstack(scores_list)
        scores_ravel = scores.ravel()
        order = scores_ravel.argsort()[::-1]
        bboxes = np.vstack(bboxes_list) / det_scale
        if self.use_kps:
            kpss = np.vstack(kpss_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 self.use_kps:
            kpss = kpss[order,:,:]
            kpss = kpss[keep,:,:]
        else:
            kpss = None
        if max_num > 0 and det.shape[0] > max_num:
            area = (det[:, 2] - det[:, 0]) * (det[:, 3] -
                                                    det[:, 1])
            img_center = img.shape[0] // 2, img.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
            bindex = np.argsort(
                values)[::-1]  # some extra weight on the centering
            bindex = bindex[0:max_num]
            det = det[bindex, :]
            if kpss is not None:
                kpss = kpss[bindex, :]
        return det, kpss

    def nms(self, dets):
        thresh = self.nms_thresh
        x1 = dets[:, 0]
        y1 = dets[:, 1]
        x2 = dets[:, 2]
        y2 = dets[:, 3]
        scores = dets[:, 4]

        areas = (x2 - x1 + 1) * (y2 - y1 + 1)
        order = scores.argsort()[::-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)

            inds = np.where(ovr <= thresh)[0]
            order = order[inds + 1]

        return keep


class FaceDetector:
    """
    This is a class for detection face bboxes and landmarks.
      
    Attributes:
        model_address (str): The model file address.
        det_size (int): Size of input frame.
        det_thresh (float): Detection threshold.
        gpu_load (boolean): Whether load model on cpu or gpu, defult=True.
        device_id (int): ID of gpu device that model should be loaded (multi-gpu case), default=0
    """
    
    def __init__(self,
                model_address="./models/scrfd_500m.onnx",
                det_size=1280,
                resize_ratio=0.8,
                det_thresh=0.5,
                gpu_load=True,
                device_id=0):
        """
        The constructor for FaceDetector class.
  
        Parameters:
            model_address (str): The model file address.
            det_size (int): Size of input frame.
            det_thresh (float): Detection threshold.
            gpu_load (boolean): Whether load model on cpu or gpu, defult=True.
            device_id (int): ID of gpu device that model should be loaded (multi-gpu case), default=0  
        """
        
        self.model_address = model_address
        self.det_size = det_size
        self.resize_ratio = resize_ratio
        print("detection resize_ratio = {}".format(self.resize_ratio))
        self.det_thresh = det_thresh
        print("detection det_thresh = {}".format(self.det_thresh))
        self.device_id = device_id
        if not os.path.exists(self.model_address):
            print("Model not found")
        
        self.detection_model = SCRFD(model_address)

        if gpu_load:
            self.detection_model.prepare(ctx_id=self.device_id, input_size=det_size, det_thresh=det_thresh)
        else:
            self.detection_model.prepare(ctx_id=-1, input_size=det_size, det_thresh=det_thresh)

    def inference(self, bgr_image, dynamic=True):
        """
        The function to inference input image from model.
  
        Parameters:
            bgr_image (numpy.ndarray): The input frame, must be a numpy ndarray in shape of (H, W, 3) and BGR format.
          
        Returns:
            faces: List of detected faces, each element is a Face object instance, Face object parameters are bbox(x1, y1, x2, y2), kps, det_score, Example: Face(bbox=bbox, kps=kps, det_score=det_score)
            inference_time: Only model inference time (second)
        """
        
        faces = []
        start_time = timeit.default_timer()
        if not dynamic:
            bboxes, kpss = self.detection_model.detect(bgr_image, input_size=(self.det_size, self.det_size), max_num=0)  # Static size inference
        else:
            dynamic_det_size = (int((bgr_image.shape[1] * self.resize_ratio) / 32) * 32, int((bgr_image.shape[0] * self.resize_ratio) / 32) * 32)  # Dynamic size input
            bboxes, kpss = self.detection_model.detect(bgr_image, input_size=dynamic_det_size, max_num=0)  # Dynamic size inference
        inference_time = timeit.default_timer() - start_time
        for i in range(bboxes.shape[0]):
            bbox = bboxes[i, 0:4]
            det_score = bboxes[i, 4]
            kps = None
            if kpss is not None:
                kps = kpss[i]
            cropped_face = bgr_image[int(bbox[1]):int(bbox[3]), int(bbox[0]):int(bbox[2]), :]
            face = Face(bbox=bbox, kps=kps, score=det_score, cropped_face=cropped_face)
            faces.append(face)
        
        return faces, inference_time, cropped_face


if __name__ == "__main__":
    detection_model = FaceDetector()
    bgr_image = cv2.imread("input/03.jpg")
    print(bgr_image.shape)
    faces, inference_time, cropped_face = detection_model.inference(bgr_image)
    bboxes = []
    confidences = []
    for face in faces:
        bboxes.append(face.bbox)
        confidences.append(face.score)
        cv2.rectangle(bgr_image, (int(face.bbox[0]), int(face.bbox[1])), (int(face.bbox[2]), int(face.bbox[3])), (144, 0, 0), 2)
    print("bboxes = {}".format(bboxes))
    print("confidences = {}".format(confidences))
    print("inference_time = {}".format(inference_time))
    cv2.imwrite("output/03.jpg", bgr_image)