OpenMMlab导出yolox模型并用onnxruntime和tensorrt推理

导出onnx文件

直接使用脚本

import torch
from mmdet.apis import init_detector, inference_detector


config_file = './configs/yolox/yolox_tiny_8xb8-300e_coco.py'
checkpoint_file = 'yolox_tiny_8x8_300e_coco_20211124_171234-b4047906.pth'
model = init_detector(config_file, checkpoint_file, device='cpu')  # or device='cuda:0'
torch.onnx.export(model, (torch.zeros(1, 3, 416, 416),), "yolox.onnx", opset_version=11)

导出的onnx结构如下：
在这里插入图片描述
输出是包含多个检测头的输出。若需要合并检测结果，需要修改脚本如下：

import torch
import cv2
import numpy as np
from mmdet.apis import init_detector, inference_detector


config_file = './configs/yolox/yolox_tiny_8xb8-300e_coco.py'
checkpoint_file = 'yolox_tiny_8x8_300e_coco_20211124_171234-b4047906.pth'
model = init_detector(config_file, checkpoint_file, device='cpu')  # or device='cuda:0'


class YOLOX(torch.nn.Module):
    def __init__(self):
        super().__init__()
        self.model = init_detector(config_file, checkpoint_file, device='cpu')
        self.class_num = 80
        self.strides = [(8, 8), (16, 16), (32, 32)]
             
    def _meshgrid(self, x, y):
        yy, xx = torch.meshgrid(y, x)
        return xx.reshape(-1), yy.reshape(-1)

    def grid_priors(self, featmap_sizes):
        multi_level_priors = []
        for i in range(len(featmap_sizes)):
            feat_h, feat_w = featmap_sizes[i]
            stride_w, stride_h = self.strides[i]
            shift_x = torch.arange(0, feat_w) * stride_w
            shift_y = torch.arange(0, feat_h) * stride_h
            shift_xx, shift_yy = self._meshgrid(shift_x, shift_y)
            stride_w = shift_xx.new_full((shift_xx.shape[0], ), stride_w)
            stride_h = shift_xx.new_full((shift_yy.shape[0], ), stride_h)
            shifts = torch.stack([shift_xx, shift_yy, stride_w, stride_h], dim=-1)       
            multi_level_priors.append(shifts)
        return multi_level_priors
    
    def bbox_decode(self, priors, bbox_preds):
        xys = (bbox_preds[..., :2] * priors[:, 2:]) + priors[:, :2]
        whs = bbox_preds[..., 2:].exp() * priors[:, 2:]
        tl_x = (xys[..., 0] - whs[..., 0] / 2)
        tl_y = (xys[..., 1] - whs[..., 1] / 2)
        br_x = (xys[..., 0] + whs[..., 0] / 2)
        br_y = (xys[..., 1] + whs[..., 1] / 2)
        decoded_bboxes = torch.stack([tl_x, tl_y, br_x, br_y], -1)
        return decoded_bboxes
        
    def forward(self, x):
        x = self.model.backbone(x)
        x = self.model.neck(x)
        pred_maps = self.model.bbox_head(x)
        
        cls_scores, bbox_preds, objectnesses = pred_maps       
        featmap_sizes = [cls_score.shape[2:] for cls_score in cls_scores]      
        mlvl_priors = self.grid_priors(featmap_sizes)
        
        flatten_cls_scores = [cls_score.permute(0, 2, 3, 1).reshape(1, -1, self.class_num) for cls_score in cls_scores]
        flatten_bbox_preds = [bbox_pred.permute(0, 2, 3, 1).reshape(1, -1, 4) for bbox_pred in bbox_preds]
        flatten_objectness = [objectness.permute(0, 2, 3, 1).reshape(1, -1) for objectness in objectnesses]
        flatten_cls_scores = torch.cat(flatten_cls_scores, dim=1).sigmoid()
        flatten_bbox_preds = torch.cat(flatten_bbox_preds, dim=1)
        flatten_objectness = torch.cat(flatten_objectness, dim=1).sigmoid()
        flatten_priors = torch.cat(mlvl_priors)
        flatten_bboxes = self.bbox_decode(flatten_priors, flatten_bbox_preds)
        
        return flatten_bboxes, flatten_objectness, flatten_cls_scores
    
    
model = YOLOX().eval()
input = torch.zeros(1, 3, 416, 416, device='cpu')
torch.onnx.export(model, input, "yolox.onnx", opset_version=11)

导出的onnx结构如下：
在这里插入图片描述

安装mmdeploy的话，可以通过下面脚本导出onnx模型。

from mmdeploy.apis import torch2onnx
from mmdeploy.backend.sdk.export_info import export2SDK


img = 'bus.jpg'
work_dir = './work_dir/onnx/yolox'
save_file = './end2end.onnx'
deploy_cfg = 'mmdeploy/configs/mmdet/detection/detection_onnxruntime_dynamic.py'
model_cfg = 'mmdetection/configs/yolox/yolox_tiny_8xb8-300e_coco.py'
model_checkpoint = 'checkpoints/yolox_tiny_8x8_300e_coco_20211124_171234-b4047906.pth'
device = 'cpu'

# 1. convert model to onnx
torch2onnx(img, work_dir, save_file, deploy_cfg, model_cfg, model_checkpoint, device)

# 2. extract pipeline info for sdk use (dump-info)
export2SDK(deploy_cfg, model_cfg, work_dir, pth=model_checkpoint, device=device)

onnx模型的结构如下：在这里插入图片描述

onnxruntime推理

手动导出的onnx模型使用onnxruntime推理：

import cv2
import numpy as np
import onnxruntime


class_names = ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light',
        'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow',
        'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',
        'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard',
        'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple',
        'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
        'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone',
        'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear',
        'hair drier', 'toothbrush'] #coco80类别     
input_shape = (416, 416)      
score_threshold = 0.2  
nms_threshold = 0.5
confidence_threshold = 0.2   


def nms(boxes, scores, score_threshold, nms_threshold):
    x1 = boxes[:, 0]
    y1 = boxes[:, 1]
    x2 = boxes[:, 2]
    y2 = boxes[:, 3]
    areas = (y2 - y1 + 1) * (x2 - x1 + 1)
    keep = []
    index = scores.argsort()[::-1] 

    while index.size > 0:
        i = index[0]
        keep.append(i)
        x11 = np.maximum(x1[i], x1[index[1:]]) 
        y11 = np.maximum(y1[i], y1[index[1:]])
        x22 = np.minimum(x2[i], x2[index[1:]])
        y22 = np.minimum(y2[i], y2[index[1:]])
        w = np.maximum(0, x22 - x11 + 1)                              
        h = np.maximum(0, y22 - y11 + 1) 
        overlaps = w * h
        ious = overlaps / (areas[i] + areas[index[1:]] - overlaps)
        idx = np.where(ious <= nms_threshold)[0]
        index = index[idx + 1]
    return keep


def filter_box(outputs): 
    outputs0, outputs1, outputs2 = outputs
    flag = outputs1 > confidence_threshold
    output0 = outputs0[flag].reshape(-1, 4)
    output1 = outputs1[flag].reshape(-1, 1)
    classes_scores = outputs2[flag].reshape(-1, 80)
    outputs = np.concatenate((output0, output1, classes_scores), axis=1)
     
    boxes = []
    scores = []
    class_ids = []
    for i in range(len(classes_scores)):
        class_id = np.argmax(classes_scores[i])
        outputs[i][4] *= classes_scores[i][class_id]
        outputs[i][5] = class_id
        if outputs[i][4] > score_threshold:
            boxes.append(outputs[i][:6])
            scores.append(outputs[i][4])
            class_ids.append(outputs[i][5])
            
    boxes = np.array(boxes)
    scores = np.array(scores)
    indices = nms(boxes, scores, score_threshold, nms_threshold) 
    output = boxes[indices]
    return output


def letterbox(im, new_shape=(416, 416), color=(114, 114, 114)):
    # Resize and pad image while meeting stride-multiple constraints
    shape = im.shape[:2]  # current shape [height, width]

    # Scale ratio (new / old)
    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
    
    # Compute padding
    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))    
    dw, dh = (new_shape[1] - new_unpad[0])/2, (new_shape[0] - new_unpad[1])/2  # wh padding 
    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
    
    if shape[::-1] != new_unpad:  # resize
        im = cv2.resize(im, new_unpad, interpolation=cv2.INTER_LINEAR)
    im = cv2.copyMakeBorder(im, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
    return im


def scale_boxes(boxes, shape):
    # Rescale boxes (xyxy) from input_shape to shape
    gain = min(input_shape[0] / shape[0], input_shape[1] / shape[1])  # gain  = old / new
    pad = (input_shape[1] - shape[1] * gain) / 2, (input_shape[0] - shape[0] * gain) / 2  # wh padding
    boxes[..., [0, 2]] -= pad[0]  # x padding
    boxes[..., [1, 3]] -= pad[1]  # y padding
    boxes[..., :4] /= gain
    boxes[..., [0, 2]] = boxes[..., [0, 2]].clip(0, shape[1])  # x1, x2
    boxes[..., [1, 3]] = boxes[..., [1, 3]].clip(0, shape[0])  # y1, y2
    return boxes


def draw(image, box_data):
    box_data = scale_boxes(box_data, image.shape)
    boxes = box_data[...,:4].astype(np.int32) 
    scores = box_data[...,4]
    classes = box_data[...,5].astype(np.int32)
   
    for box, score, cl in zip(boxes, scores, classes):
        top, left, right, bottom = box
        cv2.rectangle(image, (top, left), (right, bottom), (255, 0, 0), 1)
        cv2.putText(image, '{0} {1:.2f}'.format(class_names[cl], score), (top, left), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 0, 255), 1)


if __name__=="__main__":
    image = cv2.imread('bus.jpg')
    input = letterbox(image, input_shape)
    input = cv2.resize(image, input_shape)
    input = input[:, :, ::-1].transpose(2, 0, 1).astype(dtype=np.float32)  #BGR2RGB和HWC2CHW
    input = np.expand_dims(input, axis=0)
    
    onnx_session = onnxruntime.InferenceSession('yolox.onnx', providers=['CPUExecutionProvider'])
        
    input_name = []
    for node in onnx_session.get_inputs():
        input_name.append(node.name)

    output_name = []
    for node in onnx_session.get_outputs():
        output_name.append(node.name)

    inputs = {}
    for name in input_name:
        inputs[name] = input
        
    outputs = onnx_session.run(None, inputs)
    
    boxes = filter_box(outputs)
    draw(image, boxes)
    cv2.imwrite('result.jpg', image)

mmdeploy导出的onnx模型使用onnxruntime推理：

import cv2
import numpy as np
import onnxruntime


class_names = ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light',
        'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow',
        'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',
        'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard',
        'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple',
        'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
        'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone',
        'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear',
        'hair drier', 'toothbrush'] #coco80类别      
input_shape = (416, 416)      
confidence_threshold = 0.2


def filter_box(outputs): #删除置信度小于confidence_threshold的BOX
    flag = outputs[0][..., 4] > confidence_threshold
    boxes = outputs[0][flag] 
    class_ids = outputs[1][flag].reshape(-1, 1) 
    output = np.concatenate((boxes, class_ids), axis=1)  
    return output


def letterbox(im, new_shape=(416, 416), color=(114, 114, 114)):
    # Resize and pad image while meeting stride-multiple constraints
    shape = im.shape[:2]  # current shape [height, width]

    # Scale ratio (new / old)
    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
    
    # Compute padding
    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))    
    dw, dh = (new_shape[1] - new_unpad[0])/2, (new_shape[0] - new_unpad[1])/2  # wh padding 
    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
    
    if shape[::-1] != new_unpad:  # resize
        im = cv2.resize(im, new_unpad, interpolation=cv2.INTER_LINEAR)
    im = cv2.copyMakeBorder(im, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
    return im


def scale_boxes(input_shape, boxes, shape):
    # Rescale boxes (xyxy) from input_shape to shape
    gain = min(input_shape[0] / shape[0], input_shape[1] / shape[1])  # gain  = old / new
    pad = (input_shape[1] - shape[1] * gain) / 2, (input_shape[0] - shape[0] * gain) / 2  # wh padding

    boxes[..., [0, 2]] -= pad[0]  # x padding
    boxes[..., [1, 3]] -= pad[1]  # y padding
    boxes[..., :4] /= gain
    boxes[..., [0, 2]] = boxes[..., [0, 2]].clip(0, shape[1])  # x1, x2
    boxes[..., [1, 3]] = boxes[..., [1, 3]].clip(0, shape[0])  # y1, y2
    return boxes


def draw(image, box_data):
    box_data = scale_boxes(input_shape, box_data, image.shape)
    boxes = box_data[...,:4].astype(np.int32) 
    scores = box_data[...,4]
    classes = box_data[...,5].astype(np.int32)
   
    for box, score, cl in zip(boxes, scores, classes):
        top, left, right, bottom = box
        cv2.rectangle(image, (top, left), (right, bottom), (255, 0, 0), 1)
        cv2.putText(image, '{0} {1:.2f}'.format(class_names[cl], score), (top, left), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 0, 255), 1)


if __name__=="__main__":
    images = cv2.imread('bus.jpg')
    input = letterbox(images, input_shape)
    input = input[:, :, ::-1].transpose(2, 0, 1).astype(dtype=np.float32)  #BGR2RGB和HWC2CHW
    input = np.expand_dims(input, axis=0)
    
    onnx_session = onnxruntime.InferenceSession('../work_dir/onnx/yolox/end2end.onnx', providers=['CPUExecutionProvider'])
        
    input_name = []
    for node in onnx_session.get_inputs():
        input_name.append(node.name)

    output_name = []
    for node in onnx_session.get_outputs():
        output_name.append(node.name)

    inputs = {}
    for name in input_name:
        inputs[name] = input
        
    outputs = onnx_session.run(None, inputs)
    
    boxes = filter_box(outputs)
    draw(images, boxes)
    cv2.imwrite('result.jpg', images)

直接使用mmdeploy的api推理：

from mmdeploy.apis import inference_model


model_cfg = 'mmdetection/configs/yolox/yolox_tiny_8xb8-300e_coco.py'
deploy_cfg = 'mmdeploy/configs/mmdet/detection/detection_onnxruntime_dynamic.py'
img = 'bus.jpg'
backend_files = ['work_dir/onnx/yolox/end2end.onnx']
device = 'cpu'

result = inference_model(model_cfg, deploy_cfg, backend_files, img, device)
print(result)

或者：

from mmdeploy_runtime import Detector
import cv2

# 读取图片
img = cv2.imread('bus.jpg')

# 创建检测器
detector = Detector(model_path='work_dir/onnx/yolox', device_name='cpu')

# 执行推理
bboxes, labels, _ = detector(img)
# 使用阈值过滤推理结果，并绘制到原图中
indices = [i for i in range(len(bboxes))]
for index, bbox, label_id in zip(indices, bboxes, labels):
  [left, top, right, bottom], score = bbox[0:4].astype(int),  bbox[4]
  if score < 0.3:
      continue
  cv2.rectangle(img, (left, top), (right, bottom), (0, 255, 0))
cv2.imwrite('result.jpg', img)

导出engine文件

这里通过trtexec转换onnx文件，LZ的版本是TensorRT-8.2.1.8。

./trtexec.exe --onnx=yolox.onnx --saveEngine=yolox.engine --workspace=20480

tensorrt推理

手动导出的模型使用tensorrt推理：

import cv2
import numpy as np
import tensorrt as trt
import pycuda.autoinit 
import pycuda.driver as cuda  


class_names = ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light',
        'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow',
        'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',
        'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard',
        'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple',
        'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
        'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone',
        'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear',
        'hair drier', 'toothbrush'] #coco80类别     
input_shape = (416, 416)      
score_threshold = 0.2  
nms_threshold = 0.5
confidence_threshold = 0.2   


def nms(boxes, scores, score_threshold, nms_threshold):
    x1 = boxes[:, 0]
    y1 = boxes[:, 1]
    x2 = boxes[:, 2]
    y2 = boxes[:, 3]
    areas = (y2 - y1 + 1) * (x2 - x1 + 1)
    keep = []
    index = scores.argsort()[::-1] 

    while index.size > 0:
        i = index[0]
        keep.append(i)
        x11 = np.maximum(x1[i], x1[index[1:]]) 
        y11 = np.maximum(y1[i], y1[index[1:]])
        x22 = np.minimum(x2[i], x2[index[1:]])
        y22 = np.minimum(y2[i], y2[index[1:]])
        w = np.maximum(0, x22 - x11 + 1)                              
        h = np.maximum(0, y22 - y11 + 1) 
        overlaps = w * h
        ious = overlaps / (areas[i] + areas[index[1:]] - overlaps)
        idx = np.where(ious <= nms_threshold)[0]
        index = index[idx + 1]
    return keep


def filter_box(outputs): 
    outputs0, outputs1, outputs2 = outputs
    flag = outputs1 > confidence_threshold
    output0 = outputs0[flag].reshape(-1, 4)
    output1 = outputs1[flag].reshape(-1, 1)
    classes_scores = outputs2[flag].reshape(-1, 80)
    outputs = np.concatenate((output0, output1, classes_scores), axis=1)
     
    boxes = []
    scores = []
    class_ids = []
    for i in range(len(classes_scores)):
        class_id = np.argmax(classes_scores[i])
        outputs[i][4] *= classes_scores[i][class_id]
        outputs[i][5] = class_id
        if outputs[i][4] > score_threshold:
            boxes.append(outputs[i][:6])
            scores.append(outputs[i][4])
            class_ids.append(outputs[i][5])
            
    boxes = np.array(boxes)
    scores = np.array(scores)
    indices = nms(boxes, scores, score_threshold, nms_threshold) 
    output = boxes[indices]
    return output


def letterbox(im, new_shape=(416, 416), color=(114, 114, 114)):
    # Resize and pad image while meeting stride-multiple constraints
    shape = im.shape[:2]  # current shape [height, width]

    # Scale ratio (new / old)
    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
    
    # Compute padding
    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))    
    dw, dh = (new_shape[1] - new_unpad[0])/2, (new_shape[0] - new_unpad[1])/2  # wh padding 
    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
    
    if shape[::-1] != new_unpad:  # resize
        im = cv2.resize(im, new_unpad, interpolation=cv2.INTER_LINEAR)
    im = cv2.copyMakeBorder(im, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
    return im


def scale_boxes(boxes, shape):
    # Rescale boxes (xyxy) from input_shape to shape
    gain = min(input_shape[0] / shape[0], input_shape[1] / shape[1])  # gain  = old / new
    pad = (input_shape[1] - shape[1] * gain) / 2, (input_shape[0] - shape[0] * gain) / 2  # wh padding
    boxes[..., [0, 2]] -= pad[0]  # x padding
    boxes[..., [1, 3]] -= pad[1]  # y padding
    boxes[..., :4] /= gain
    boxes[..., [0, 2]] = boxes[..., [0, 2]].clip(0, shape[1])  # x1, x2
    boxes[..., [1, 3]] = boxes[..., [1, 3]].clip(0, shape[0])  # y1, y2
    return boxes


def draw(image, box_data):
    box_data = scale_boxes(box_data, image.shape)
    boxes = box_data[...,:4].astype(np.int32) 
    scores = box_data[...,4]
    classes = box_data[...,5].astype(np.int32)
   
    for box, score, cl in zip(boxes, scores, classes):
        top, left, right, bottom = box
        cv2.rectangle(image, (top, left), (right, bottom), (255, 0, 0), 1)
        cv2.putText(image, '{0} {1:.2f}'.format(class_names[cl], score), (top, left), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 0, 255), 1)


if __name__=="__main__":
    logger = trt.Logger(trt.Logger.WARNING)
    with open("yolox.engine", "rb") as f, trt.Runtime(logger) as runtime:
        engine = runtime.deserialize_cuda_engine(f.read())
    context = engine.create_execution_context()
    h_input = cuda.pagelocked_empty(trt.volume(context.get_binding_shape(0)), dtype=np.float32)
    h_output0 = cuda.pagelocked_empty(trt.volume(context.get_binding_shape(1)), dtype=np.float32)
    h_output1 = cuda.pagelocked_empty(trt.volume(context.get_binding_shape(2)), dtype=np.float32)
    h_output2 = cuda.pagelocked_empty(trt.volume(context.get_binding_shape(3)), dtype=np.float32)
    d_input = cuda.mem_alloc(h_input.nbytes)
    d_output0 = cuda.mem_alloc(h_output0.nbytes)
    d_output1 = cuda.mem_alloc(h_output1.nbytes)
    d_output2 = cuda.mem_alloc(h_output2.nbytes)
    stream = cuda.Stream()
    
    image = cv2.imread('bus.jpg')
    input = letterbox(image, input_shape)
    input = input[:, :, ::-1].transpose(2, 0, 1).astype(dtype=np.float32)  #BGR2RGB和HWC2CHW
    input = np.expand_dims(input, axis=0)  
    np.copyto(h_input, input.ravel())

    with engine.create_execution_context() as context:
        cuda.memcpy_htod_async(d_input, h_input, stream)
        context.execute_async_v2(bindings=[int(d_input), int(d_output0), int(d_output1), int(d_output2)], stream_handle=stream.handle)
        cuda.memcpy_dtoh_async(h_output0, d_output0, stream)
        cuda.memcpy_dtoh_async(h_output1, d_output1, stream)
        cuda.memcpy_dtoh_async(h_output2, d_output2, stream)
        stream.synchronize()  
        h_output = []
        h_output.append(h_output2.reshape(1, 3549, 4))
        h_output.append(h_output1.reshape(1, 3549))
        h_output.append(h_output0.reshape(1, 3549, 80))
        boxes = filter_box(h_output)
        draw(image, boxes)
        cv2.imwrite('result.jpg', image)

使用mmdeploy的api推理：

from mmdeploy.apis import torch2onnx
from mmdeploy.backend.sdk.export_info import export2SDK


img = 'bus.jpg'
work_dir = './work_dir/onnx/yolox'
save_file = './end2end.onnx'
deploy_cfg = 'mmdeploy/configs/mmdet/detection/detection_onnxruntime_dynamic.py'
model_cfg = 'mmdetection/configs/yolox/yolox_tiny_8xb8-300e_coco.py'
model_checkpoint = 'checkpoints/yolox_tiny_8x8_300e_coco_20211124_171234-b4047906.pth'
device = 'cpu'

# 1. convert model to onnx
torch2onnx(img, work_dir, save_file, deploy_cfg, model_cfg, model_checkpoint, device)

# 2. extract pipeline info for sdk use (dump-info)
export2SDK(deploy_cfg, model_cfg, work_dir, pth=model_checkpoint, device=device)

或者

from mmdeploy_runtime import Detector
import cv2

# 读取图片
img = cv2.imread('bus.jpg')

# 创建检测器
detector = Detector(model_path='work_dir/trt/yolox', device_name='cuda')

# 执行推理
bboxes, labels, _ = detector(img)
# 使用阈值过滤推理结果，并绘制到原图中
indices = [i for i in range(len(bboxes))]
for index, bbox, label_id in zip(indices, bboxes, labels):
  [left, top, right, bottom], score = bbox[0:4].astype(int),  bbox[4]
  if score < 0.3:
      continue
  cv2.rectangle(img, (left, top), (right, bottom), (0, 255, 0))
cv2.imwrite('result.jpg', img)