# EASE : Use POT to Quantize single-human-pose-estimation-0001 Public Model
###### tags: `EASE` `POT` `2022.1`
## Use OpenVINO dockerhub image
```
docker run -it -v ~/Downloads:/mnt -u root --rm openvino/ubuntu20_dev:latest
```
## Run Accuracy Checker and POT
In ubuntu20_dev docker image,
#### 0. Download COCO 2017 trainval dataset and annotation
```=bash
cd /home/openvino
apt update
apt install unzip
mkdir coco_dataset
cd coco_dataset
curl http://images.cocodataset.org/zips/val2017.zip -o val2017.zip
unzip val2017.zip
curl http://images.cocodataset.org/annotations/annotations_trainval2017.zip -o trainval_2017.zip
unzip trainval_2017.zip
```
##### coco_dataset content
```
coco_dataset/
|--annotations/
|-- captions_train2017.json
|-- captions_val2017.json
|-- instances_train2017.json
|-- instances_val2017.json
|-- person_keypoints_train2017.json
`-- person_keypoints_val2017.json
|-- val2017/
|-- 000000042102.jpg
|-- 000000060102.jpg
|-- 000000245102.jpg
...
`-- 000000364102.jpg
```
#### 1. Download single-human-pose-estimation-0001
omz_downloader --name single-human-pose-estimation-0001 -o /home/openvino/openvino_models
#### 2. Convert single-human-pose-estimation-0001 to IR
omz_converter --name single-human-pose-estimation-0001 -d /home/openvino/openvino_models
#### 3. Run Accuracy Checker on single-human-pose-estimation-0001
pip install pycocotools
accuracy_check -c single-human-pose-estimation-0001-int8.yml -m openvino_models/public/single-human-pose-estimation-0001/FP32/
#### 4. Run POT on single-human-pose-estimation-0001
pot -c single-human-pose-estimation-0001-int8.json -e
#### 5. Copy single-human-pose-estimation-0001 FP16-INT8 IR
```
mkdir /home/openvino/openvino_models/public/single-human-pose-estimation-0001/FP16-INT8/
cp -ar results/single-human-pose-estimation-0001*/*/optimized/* /home/openvino/openvino_models/public/single-human-pose-estimation-0001/FP16-INT8/
```
## Reference
https://raw.githubusercontent.com/openvinotoolkit/open_model_zoo/master/models/public/single-human-pose-estimation-0001/accuracy-check.yml
### single-human-pose-estimation-0001-int8.yml
```
models:
- name: single-human-pose-estimation-0001
launchers:
- framework: openvino
device: CPU
adapter:
type: single_human_pose_estimation
datasets:
- name: ms_coco_single_keypoints
data_source: /home/openvino/coco_dataset/val2017
annotation_conversion:
converter: mscoco_single_keypoints
annotation_file: /home/openvino/coco_dataset/annotations/person_keypoints_val2017.json
annotation: mscoco_single_keypoints.pickle
dataset_meta: mscoco_single_keypoints.json
preprocessing:
- type: transformed_crop_with_auto_scale
dst_height: 384
dst_width: 288
stride: 8
metrics:
- name: AP
type: coco_orig_keypoints_precision
reference: 0.6904
```
### single-human-pose-estimation-0001.json
```
{
"model": {
"model_name": "single-human-pose-estimation-0001",
"model": "/home/openvino/openvino_models/public/single-human-pose-estimation-0001/FP16/single-human-pose-estimation-0001.xml",
"weights": "/home/openvino/openvino_models/public/single-human-pose-estimation-0001/FP16/single-human-pose-estimation-0001.bin"
},
"engine": {
"config": "/home/openvino/single-human-pose-estimation-0001-int8.yml"
},
"compression": {
"algorithms": [
{
"name": "DefaultQuantization",
"params": {
"preset": "performance",
"stat_subset_size": 100
}
}
]
}
}
```
### accuracy_checker log
```
accuracy_check -c single-human-pose-estimation-0001-int8.yml -m openvino_models/public/single-human-pose-estimation-0001/FP32/single-human-pose-estimation-0001.xml
Processing info:
model: single-human-pose-estimation-0001
launcher: openvino
device: CPU
dataset: ms_coco_single_keypoints
OpenCV version: 4.5.5
Annotation for ms_coco_single_keypoints dataset will be loaded from mscoco_single_keypoints.pickle
Loaded dataset info:
Dataset name: ms_coco_single_keypoints
Accuracy Checker version 0.9.3
Dataset size 6352
Conversion parameters:
converter: mscoco_single_keypoints
annotation_file: /home/openvino/coco_dataset/annotations/person_keypoints_val2017.json
ms_coco_single_keypoints dataset metadata will be loaded from mscoco_single_keypoints.json
IE version: 2022.1.0-7019-cdb9bec7210-releases/2022/1
Loaded CPU plugin version:
CPU - openvino_intel_cpu_plugin: 2022.1.2022.1.0-7019-cdb9bec7210-releases/2022/1
Found model openvino_models/public/single-human-pose-estimation-0001/FP32/single-human-pose-estimation-0001.xml
Found weights openvino_models/public/single-human-pose-estimation-0001/FP32/single-human-pose-estimation-0001.bin
Input info:
Node name: data
Tensor names: data
precision: f32
shape: (1, 3, 384, 288)
Output info
Node name: heatmaps/sink_port_0
Tensor names: heatmaps
precision: f32
shape: (1, 17, 48, 36)
6352 objects processed in 774.886 seconds
loading annotations into memory...
Done (t=0.20s)
creating index...
index created!
Loading and preparing results...
DONE (t=0.31s)
creating index...
index created!
Running per image evaluation...
Evaluate annotation type *keypoints*
DONE (t=3.88s).
Accumulating evaluation results...
DONE (t=0.08s).
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets= 20 ] = 0.690
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets= 20 ] = 0.923
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets= 20 ] = 0.779
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.662
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.732
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 20 ] = 0.726
Average Recall (AR) @[ IoU=0.50 | area= all | maxDets= 20 ] = 0.931
Average Recall (AR) @[ IoU=0.75 | area= all | maxDets= 20 ] = 0.808
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.692
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.776
MSCOCOorigBaseMetric.compute_precision_recall: returning [[0.690453523582747, 0.9232926843990009, 0.7791885958151511, 0.6617599597063265, 0.731706173047956], [0.7259603274559194, 0.9310453400503779, 0.8082493702770781, 0.6921606118546846, 0.7761426978818284]]
AP: 69.05% [FAILED: abs error = 0.005352 | relative error = 7.753e-05]
```
# POT logs
```
pot -c single-human-pose-estimation-0001-int8.json
INFO:openvino.tools.pot.app.run:Output log dir: ./results/single-human-pose-estimation-0001_DefaultQuantization/2022-04-27_14-35-07
INFO:openvino.tools.pot.app.run:Creating pipeline:
Algorithm: DefaultQuantization
Parameters:
preset : performance
stat_subset_size : 100
target_device : ANY
model_type : None
dump_intermediate_model : False
inplace_statistics : True
exec_log_dir : ./results/single-human-pose-estimation-0001_DefaultQuantization/2022-04-27_14-35-07
===========================================================================
IE version: 2022.1.0-7019-cdb9bec7210-releases/2022/1
Loaded CPU plugin version:
CPU - openvino_intel_cpu_plugin: 2022.1.2022.1.0-7019-cdb9bec7210-releases/2022/1
Annotation for ms_coco_single_keypoints dataset will be loaded from mscoco_single_keypoints.pickle
Loaded dataset info:
Dataset name: ms_coco_single_keypoints
Accuracy Checker version 0.9.3
Dataset size 6352
Conversion parameters:
converter: mscoco_single_keypoints
annotation_file: /home/openvino/coco_dataset/annotations/person_keypoints_val2017.json
ms_coco_single_keypoints dataset metadata will be loaded from mscoco_single_keypoints.json
INFO:openvino.tools.pot.pipeline.pipeline:Inference Engine version: 2022.1.0-7019-cdb9bec7210-releases/2022/1
INFO:openvino.tools.pot.pipeline.pipeline:Model Optimizer version: 2022.1.0-7019-cdb9bec7210-releases/2022/1
INFO:openvino.tools.pot.pipeline.pipeline:Post-Training Optimization Tool version: 2022.1.0-7019-cdb9bec7210-releases/2022/1
INFO:openvino.tools.pot.statistics.collector:Start computing statistics for algorithms : DefaultQuantization
INFO:openvino.tools.pot.statistics.collector:Computing statistics finished
INFO:openvino.tools.pot.pipeline.pipeline:Start algorithm: DefaultQuantization
INFO:openvino.tools.pot.algorithms.quantization.default.algorithm:Start computing statistics for algorithm : ActivationChannelAlignment
INFO:openvino.tools.pot.algorithms.quantization.default.algorithm:Computing statistics finished
INFO:openvino.tools.pot.algorithms.quantization.default.algorithm:Start computing statistics for algorithms : MinMaxQuantization,FastBiasCorrection
INFO:openvino.tools.pot.algorithms.quantization.default.algorithm:Computing statistics finished
INFO:openvino.tools.pot.pipeline.pipeline:Finished: DefaultQuantization
```
Sign in with Wallet
Connect another wallet