mirror of
https://github.com/open-mmlab/mmdeploy.git
synced 2025-01-14 08:09:43 +08:00
* Fix include and lib paths for onnxruntime. * Fixes for SSD export test * Add onnx2openvino and OpenVINODetector. Test models: ssd, retinanet, fcos, fsaf. * Add support for two-stage models: faster_rcnn, cascade_rcnn * Add doc * Add strip_doc_string for openvino. * Fix openvino preprocess. * Add OpenVINO to test_wrapper.py. * Fix * Add openvino_execute. * Removed preprocessing. * Fix onnxruntime cmake. * Rewrote postprocessing and forward, added docstrings and fixes. * Added device type change to OpenVINOWrapper. * Update forward_of_single_roi_extractor_dynamic_openvino and fix doc. * Update docs. * Add OpenVINODetector and onn2openvino tests. * Add input_info to onnx2openvino. * Add TestOpenVINOExporter and test_single_roi_extractor. * Moved get_input_shape_from_cfg to openvino_utils.py and added test. * Added test_cascade_roi_head. * Add backend.check_env() to tests. * Add OpenVINO to get_rewrite_outputs and to some tests in test_mmdet_models. * Moved test_single_roi_extractor to test_mmdet_models. * Removed TestOpenVINOExporter.
1.9 KiB
1.9 KiB
OpenVINO Support
This tutorial is based on Linux systems like Ubuntu-18.04.
Installation
- Install OpenVINO.
- Install MMdeploy following the instructions.
Usage
Example:
python tools/deploy.py \
configs/mmdet/single-stage/single-stage_openvino_dynamic.py \
/mmdetection_dir/mmdetection/configs/ssd/ssd300_coco.py \
/tmp/snapshots/ssd300_coco_20210803_015428-d231a06e.pth \
tests/data/tiger.jpeg \
--work-dir ../deploy_result \
--device cpu \
--log-level INFO \
List of supported models exportable to OpenVINO from MMDetection
The table below lists the models that are guaranteed to be exportable to OpenVINO from MMDetection.
Model | Config | Dynamic Shape |
---|---|---|
FCOS | configs/fcos/fcos_r50_caffe_fpn_gn-head_4x4_1x_coco.py |
Y |
FSAF | configs/fsaf/fsaf_r50_fpn_1x_coco.py |
Y |
RetinaNet | configs/retinanet/retinanet_r50_fpn_1x_coco.py |
Y |
SSD | configs/ssd/ssd300_coco.py |
Y |
Faster R-CNN | configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py |
Y |
Cascade R-CNN | configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco.py |
Y |
Notes:
- For faster work in OpenVINO in the Faster-RCNN, Cascade-RCNN the RoiAlign operation is replaced with the ExperimentalDetectronROIFeatureExtractor operation in the ONNX graph.