MMDetection Deployment

MMDetection aka mmdet is an open source object detection toolbox based on PyTorch. It is a part of the OpenMMLab project.


Install mmdet

Please follow the installation guide to install mmdet.

Install mmdeploy

There are several methods to install mmdeploy, among which you can choose an appropriate one according to your target platform and device.

Method I: Install precompiled package

You can refer to get_started

Method II: Build using scripts

If your target platform is Ubuntu 18.04 or later version, we encourage you to run scripts. For example, the following commands install mmdeploy as well as inference engine - ONNX Runtime.

git clone --recursive -b main
cd mmdeploy
python3 tools/scripts/ $(nproc)
export PYTHONPATH=$(pwd)/build/lib:$PYTHONPATH
export LD_LIBRARY_PATH=$(pwd)/../mmdeploy-dep/onnxruntime-linux-x64-1.8.1/lib/:$LD_LIBRARY_PATH

Method III: Build from source

If neither I nor II meets your requirements, building mmdeploy from source is the last option.

Convert model

You can use tools/ to convert mmdet models to the specified backend models. Its detailed usage can be learned from here.

The command below shows an example about converting Faster R-CNN model to onnx model that can be inferred by ONNX Runtime.

cd mmdeploy
# download faster r-cnn model from mmdet model zoo
mim download mmdet --config faster-rcnn_r50_fpn_1x_coco --dest .
# convert mmdet model to onnxruntime model with dynamic shape
python tools/ \
    configs/mmdet/detection/ \ \
    faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth \
    demo/resources/det.jpg \
    --work-dir mmdeploy_models/mmdet/ort \
    --device cpu \
    --show \

It is crucial to specify the correct deployment config during model conversion. We’ve already provided builtin deployment config files of all supported backends for mmdetection, under which the config file path follows the pattern:

{task}/{task}_{backend}-{precision}_{static | dynamic}_{shape}.py
  • {task}: task in mmdetection.

    There are two of them. One is detection and the other is instance-seg, indicating instance segmentation.

    mmdet models like RetinaNet, Faster R-CNN and DETR and so on belongs to detection task. While Mask R-CNN is one of instance-seg models. You can find more of them in chapter Supported models.

    DO REMEMBER TO USE detection/detection_*.py deployment config file when trying to convert detection models and use instance-seg/instance-seg_*.py to deploy instance segmentation models.

  • {backend}: inference backend, such as onnxruntime, tensorrt, pplnn, ncnn, openvino, coreml etc.

  • {precision}: fp16, int8. When it’s empty, it means fp32

  • {static | dynamic}: static shape or dynamic shape

  • {shape}: input shape or shape range of a model

Therefore, in the above example, you can also convert faster r-cnn to other backend models by changing the deployment config file to others, e.g., converting to tensorrt-fp16 model by


When converting mmdet models to tensorrt models, –device should be set to “cuda”

Model specification

Before moving on to model inference chapter, let’s know more about the converted model structure which is very important for model inference.

The converted model locates in the working directory like mmdeploy_models/mmdet/ort in the previous example. It includes:

├── deploy.json
├── detail.json
├── end2end.onnx
└── pipeline.json

in which,

  • end2end.onnx: backend model which can be inferred by ONNX Runtime

  • *.json: the necessary information for mmdeploy SDK

The whole package mmdeploy_models/mmdet/ort is defined as mmdeploy SDK model, i.e., mmdeploy SDK model includes both backend model and inference meta information.

Model inference

Backend model inference

Take the previous converted end2end.onnx model as an example, you can use the following code to inference the model and visualize the results.

from mmdeploy.apis.utils import build_task_processor
from mmdeploy.utils import get_input_shape, load_config
import torch

deploy_cfg = 'configs/mmdet/detection/'
model_cfg = './'
device = 'cpu'
backend_model = ['./mmdeploy_models/mmdet/ort/end2end.onnx']
image = './demo/resources/det.jpg'

# read deploy_cfg and model_cfg
deploy_cfg, model_cfg = load_config(deploy_cfg, model_cfg)

# build task and backend model
task_processor = build_task_processor(model_cfg, deploy_cfg, device)
model = task_processor.build_backend_model(backend_model)

# process input image
input_shape = get_input_shape(deploy_cfg)
model_inputs, _ = task_processor.create_input(image, input_shape)

# do model inference
with torch.no_grad():
    result = model.test_step(model_inputs)

# visualize results

SDK model inference

You can also perform SDK model inference like following,

from mmdeploy_runtime import Detector
import cv2

img = cv2.imread('./demo/resources/det.jpg')

# create a detector
detector = Detector(model_path='./mmdeploy_models/mmdet/ort', device_name='cpu', device_id=0)
# perform inference
bboxes, labels, masks = detector(img)

# visualize inference result
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:

  cv2.rectangle(img, (left, top), (right, bottom), (0, 255, 0))

cv2.imwrite('output_detection.png', img)

Besides python API, mmdeploy SDK also provides other FFI (Foreign Function Interface), such as C, C++, C#, Java and so on. You can learn their usage from demos.

Supported models

Model Task OnnxRuntime TensorRT ncnn PPLNN OpenVINO
ATSS Object Detection Y Y N N Y
FCOS Object Detection Y Y Y N Y
FoveaBox Object Detection Y N N N Y
FSAF Object Detection Y Y Y Y Y
RetinaNet Object Detection Y Y Y Y Y
SSD Object Detection Y Y Y N Y
VFNet Object Detection N N N N Y
YOLOv3 Object Detection Y Y Y N Y
YOLOX Object Detection Y Y Y N Y
Cascade R-CNN Object Detection Y Y N Y Y
Faster R-CNN Object Detection Y Y Y Y Y
Faster R-CNN + DCN Object Detection Y Y Y Y Y
GFL Object Detection Y Y N ? Y
RepPoints Object Detection N Y N ? Y
DETR* Object Detection Y Y N ? Y
Deformable DETR* Object Detection Y Y N ? Y
Conditional DETR* Object Detection Y Y N ? Y
DAB-DETR* Object Detection Y Y N ? Y
DINO* Object Detection Y Y N ? Y
CenterNet Object Detection Y Y N ? Y
RTMDet Object Detection Y Y N ? Y
Cascade Mask R-CNN Instance Segmentation Y Y N N Y
HTC Instance Segmentation Y Y N ? Y
Mask R-CNN Instance Segmentation Y Y N N Y
Swin Transformer Instance Segmentation Y Y N N Y
SOLO Instance Segmentation Y N N N Y
SOLOv2 Instance Segmentation Y N N N Y
CondInst Instance Segmentation Y Y N N N
Panoptic FPN Panoptic Segmentation Y Y N N N
MaskFormer Panoptic Segmentation Y Y N N N
Mask2Former* Panoptic Segmentation Y Y N N N


  • For transformer based models, strongly suggest use TensorRT>=8.4.

  • Mask2Former should use TensorRT>=8.6.1 for dynamic shape inference.

  • DETR-like models do not support multi-batch inference.

Read the Docs v: v1.3.1
On Read the Docs
Project Home

Free document hosting provided by Read the Docs.