Target of the V1.1.0

This is the formal release V1.1.0 of SNNF(Sunplus Neural Network Framework).

On the basis of v1.0.0, v1.1.0 mainly has the following updates：

Add some NN module.
- genderage: gender and age
  - Det_10g + genderage
  - LightFace + genderage
Based on the current main framework of SNNF, provide a porting tool so that users can easily add a new model to the Customized zone using the official generated nbg code as the source. The initial version of this tool is currently available, and there will be further updates in the next version.
Fix some known issues, such as:
OCR have a probability of crashing.
HumanAttr have a probability of crashing.
When multiple models are connected in series, if the last model comes out, the result may have an unfinished flag fin=false.
Standardize post-process code.
Organize samples and unittest codes:
samples is a sample code for snnf command sample which is just for snnf_run.sh.
unittest/modelTest is just for a buildin model's test sample.
unittest/pluginTest is just for the plugin model's test sample.
unittest/sequntialTest is the sample for the pipeline in which several models are sequentially linked together in series.
Add NN filter, which can facilitate users to do their own private data processing at the application layer. Users only need to insert the filter directly into the corresponding position of the pipeline when creating the model flow, and then perform private and correct processing at the most appropriate location. The advantage of this is that there is no need to write a dedicated module within SNNF for this private processing, only the official interface of SNNF needs to be called. Of course, SNNF provides some commonly used private processing filters internally, such as label filtering. For example, when Object Detect+HumanAttr, the label results of Object Detect can be filtered, and only the person can be passed to HumanAttr. If the requirements of the application layer happen to match this usage, there is no need to implement it yourself at the application layer. Simply call the methods provided by SNNF.
Provide users with a mAP test sample to compare the precision of the C3V platform's model running results with the precision of the model's official PC running results. The default test model is yolov8s object detection. Users can refer to this mAP sample to rewrite the model they need to test mAP values for.
Improve the toString() function for each model's result, and support the detection result toString JSON format.
Update some make config.
In addition to cross compilation on the server, we also provide compilation methods on the C3V platform.
If it is cross compilation on the server, please confirm the config CROSS_COMPILE := ON in makefile_config.mk .
If compilation is required on the C3V platform, please confirm the config CROSS_COMPILE := OFF in makefile_config.mk .
Users can choose the model plugin in makefile_config.mk that needs to be compiled according to their own needs, which can effectively reduce the code size of specific projects and shorten compilation time.

# ------------------------------------------------------------------------
# define which PLUGIN to build
  NN_PLUGIN_ENABLE_YOLOV5S_V2                 := yes

  NN_PLUGIN_ENABLE_RTMDETS                    := no
  NN_PLUGIN_ENABLE_GEDERAGE                   := no
  NN_PLUGIN_ENABLE_YOLOV8S_DETECTION          := no
  NN_PLUGIN_ENABLE_YOLOV8S_OBB                := no
  NN_PLUGIN_ENABLE_YOLOV8S_POSE               := no
  NN_PLUGIN_ENABLE_YOLOV8S_SEGMENT            := no
  NN_PLUGIN_ENABLE_YOLOV10S_DETECTION         := no

  NN_PLUGIN_ENABLE_YOLOV8N_CCPD               := no
  NN_PLUGIN_ENABLE_YOLOV8N_OBB_OPTI           := no
  NN_PLUGIN_ENABLE_YOLOV8N_POSE_OPTI          := no
  NN_PLUGIN_ENABLE_YOLOV8N_SEGMENT_OPTI       := no
  NN_PLUGIN_ENABLE_YOLOV8N_DETECTION_OPTI     := no
  NN_PLUGIN_ENABLE_YOLOV8S_DETECTION_OPTI     := no

  NN_PLUGIN_ENABLE_BOTSORT_TRACK              := no

Adjust the access of certain model header files and provide more necessary header files for the application layer to use.
Change the bbox value type from uint to float. Although this change does not have a substantial impact on the rendering of the results, it can make the precision value more accurate when calculating mAP.
In previous versions, assist modules such as imageWriter/videoWriter/JsonWriter only supported the default filename. Starting from this version, we will support filename config.
Also provide release folder so that new users can easily preview or evaluate our SNNF demo without spending too much time on environment setup and code compilation.

Resource

Please get the V1.1.0 release resource here.

Before starting any work, please carefully read the instruction files such as readme.md in the document directory.

Usage of V1.1.0

How to verify Official Demos

You can use the script we provide to start as follows：

/SNNF/release # ./snnf_run.sh
Usage: ./bin/snnf_nnsample [-m|-s|-a|-h] [-i|-v|-o|option]
        Version: 1.1.0_
        Time:
        [-m,--model <model>] run a single model
                <model>:Age                        Det10g                     HumanAttr
                        HumanFilter                LightFace                  OcrCls
                        OcrDet                     OcrRec                     VehicleAttr
                        VehicleFilter              W600kR50                   YoloV8sOdMap
                        Yolov5sDetection           Yolov5sV1                  Yolov8nClassify
                        Yolov8sClassify            stgcn                      Yolov5sV2
                        BotSortTrackStgcn          GenderAge                  Rtmdets
                        YoloV10sDetection          YoloV8nCcpdOpti            YoloV8nDetectionBaseOpti
                        YoloV8nDetectionOpti       YoloV8nObbOpti             YoloV8nPoseOpti
                        YoloV8nSegmentOpti         YoloV8sDetection           YoloV8sDetectionBaseOpti
                        YoloV8sDetectionOpti       YoloV8sObb                 YoloV8sPose
                        YoloV8sSegment
                example:./bin/snnf_nnsample -m Yolov5sDetection
                        ./bin/snnf_nnsample --model HumanAttr

        [-s,--sequential <model1,model2,...>] run sequential models
                <models>:Yolov5sDetection,HumanFilter,HumanAttr
                        LightFace,Age
                        OcrDet,OcrRec
                        OcrDet,OcrCls,OcrRec
                        YoloV8nCcpdOpti,OcrRec
                        Det10g,W600kR50
                        YoloV8sDetection,VehicleFilter,VehicleAttr
                        YoloV8nDetectionOpti,BotSortTrack
                        YoloV8nPoseOpti,BotSortTrackStgcn
                        YoloV8nCcpdOpti,OcrRec
                        YoloV8nCcpdOpti,OcrCls,OcrRec
                example:./bin/snnf_nnsample -s Yolov5sDetection,HumanFilter,HumanAttr
                        ./bin/snnf_nnsample --sequential ocrDet,ocrCls,ocrRec
                        ./bin/snnf_nnsample -s YoloV8nCcpdOpti,OcrRec,imageWriter
                        ./bin/snnf_nnsample -s YoloV8sDetectionOpti,BotSortTrack,videoWriter -v resource/video/humanCount.mp4
                        ./bin/snnf_nnsample -s YoloV8nPoseOpti,BotSortTrackStgcn,videoWriter -v resource/video/person-falling.mp4

        [-i,--image file] set image file to nn detection.
                <file>: file name
                [-c | option]: test count, this parameter is only match with -i
                example:./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr -i filename -c testCount
                        ./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr --image filename -c testCount

        [-v,--video file] set video file to nn detection.
                <file>: file name
                example:./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr -v filename
                        ./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr --video filename

        [-o,--output file] specify the output file name for saving results.
                <file>: file name with extension (e.g., output.jpg, output.json, output.mp4)
                This parameter must be used in conjunction with imageWriter, jsonWriter, or videoWriter.
                example:./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr,imageWriter -i filename -o output.jpg
                        ./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr,jsonWriter -i filename -o output.json
                        ./bin/snnf_nnsample -s Yolov5sDetection,HumanAttr,videoWriter -v filename -o output.mp4
                        
        [-a,--all] run all model testing

        assist tools: imageWriter videoWriter jsonWriter BotSortTrack HumanFilter VehicleFilter

Release folder structure

bin: nnf_nnsample. Prebuild sample programs that can run on the c3v Linux platform.
include：header file of NN framework SDK.
lib：libraries of NN framework SDK.
resource
- config: some config files for features.
- font: ttf file for plotting sample.
- image: image files used for test.
- model: models to be used in the sample program.
- video: video files used for test.
samples：example code for using NN framework.
snnf_run.sh：executable script for running sample code.
thirdparty: just as its name implies.

How to run NN framework sample

Copy the release foler to C3V Linux.

/SNNF/release # ls -alh
total 36
drwxr-xr-x  8 xxx B400 4096 Sep 30 14:23 ./
drwxr-xr-x 15 xxx B400 4096 Sep 30 14:15 ../
drwxr-xr-x  2 xxx B400 4096 Sep 30 14:15 bin/
drwxr-xr-x  6 xxx B400 4096 Sep 30 14:15 include/
drwxr-xr-x  3 xxx B400 4096 Sep 30 14:15 lib/
drwxr-xr-x  7 xxx B400 4096 Sep 30 14:15 resource/
drwxr-xr-x  4 xxx B400 4096 Sep 30 14:15 samples/
-rwxr-xr-x  1 xxx B400  262 Sep 30 14:15 snnf_run.sh*
drwxr-xr-x  7 xxx B400 4096 Sep 30 14:15 thirdparty/

Run nnf_run.sh to run the NNF sample.

a. One-time input

./snnf_run.sh -m YoloV8sDetection

#./snnf_run.sh -m YoloV8sDetection
1727625905242|7fad06c020|T|common: [app]YoloV8sDetection in
1727625905261|7fad06c020|I|common: [nn]create model from pluginName: YoloV8sDetection takes: 17
1727625905722|7f96fdf0e0|I|common: [nn]picked: 5
1727625905722|7f96fdf0e0|T|common: [app]GeneralModelOutputListener detect from resource/image/objectDetect.jpg, the result: (box:   35  196  158  403) --> label: 0(person), confidence: 0.89, fin: false
1727625905722|7f96fdf0e0|T|common: [app]GeneralModelOutputListener detect from resource/image/objectDetect.jpg, the result: (box:  526  185  112  392) --> label: 0(person), confidence: 0.87, fin: false
1727625905722|7f96fdf0e0|T|common: [app]GeneralModelOutputListener detect from resource/image/objectDetect.jpg, the result: (box:  173  203   97  360) --> label: 0(person), confidence: 0.86, fin: false
1727625905722|7f96fdf0e0|T|common: [app]GeneralModelOutputListener detect from resource/image/objectDetect.jpg, the result: (box:    0  321   44  253) --> label: 0(person), confidence: 0.75, fin: false
1727625905722|7f96fdf0e0|T|common: [app]GeneralModelOutputListener detect from resource/image/objectDetect.jpg, the result: (box:   12   66  625  403) --> label: 5(bus), confidence: 0.74, fin: true
1727625905781|7fad06c020|T|common: [app]YoloV8sDetection out, retVal: 0

b. Read input from the image file

./snnf_run.sh -m YoloV8nDetectionOpti -i resource/image/person640x640.jpg

# ./snnf_run.sh -m YoloV8nDetectionOpti -i resource/image/person640x640.jpg
1727626580087|7fbd7de020|T|common: [app]YoloV8nDetectionOpti in
1727626582152|7fbd7de020|I|common: [nn]create model from pluginName: YoloV8nDetectionOpti takes: 2063
1727626582574|7fa569a0e0|T|common: [app]GeneralModelOutputListener detect from resource/image/person640x640.jpg, the result: (box:    0   19  614  619) --> label: 0(person), confidence: 0.87, fin: true
1727626582622|7fbd7de020|T|common: [app]YoloV8nDetectionOpti out, retVal: 0

c. Read inputs from the video file.

./snnf_run.sh -m YoloV8nDetectionOpti -v resource/video/humanCount.mp4

# ./snnf_run.sh -m YoloV8nDetectionOpti -v resource/video/humanCount.mp4
1727626644381|7faf81e020|T|common: [app]streaming in
1727626645952|7faf81e020|I|common: [nn]create model from pluginName: YoloV8nDetectionOpti takes: 1388
1727626646044|7f951d80e0|T|common: [app]streaming test: runner func in
1727626646304|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  671  129  307  873) --> label: 0(person), confidence: 0.90, fin: false
1727626646305|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:    0  364  268  461) --> label: 7(truck), confidence: 0.38, fin: true
1727626646375|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  672  127  325  873) --> label: 0(person), confidence: 0.90, fin: false
1727626646375|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:    0  364  267  461) --> label: 7(truck), confidence: 0.39, fin: true
1727626646440|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  671  125  327  878) --> label: 0(person), confidence: 0.90, fin: false
1727626646440|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:    0  364  267  460) --> label: 7(truck), confidence: 0.41, fin: true
1727626646498|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  673  125  323  878) --> label: 0(person), confidence: 0.90, fin: false
......
1727626682667|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  977  159  330  894) --> label: 0(person), confidence: 0.91, fin: false
1727626682667|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  668  262  270  708) --> label: 0(person), confidence: 0.86, fin: false
1727626682667|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:    1  363  267  457) --> label: 7(truck), confidence: 0.35, fin: false
1727626682667|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  465   26 1454 1003) --> label: 6(train), confidence: 0.32, fin: false
1727626682667|7f959e80e0|T|common: [app]GeneralModelOutputListener detect from , the result: (box:  847  565  173  403) --> label: 0(person), confidence: 0.26, fin: true

1727626686753|7faf81e020|T|common: [app]q to quit

d. Sequential models

./snnf_run.sh -s Yolov5sDetection,HumanAttr

# ./snnf_run.sh -s Yolov5sDetection,HumanAttr
1727627713332|7fa3285020|T|common: [app]sequential in
1727627713501|7fa3285020|T|common: [app]input image name: resource/image/person.jpg
1727627713833|7f937c50e0|T|common: [nn]detectedInfos: 1
1727627713833|7f937c50e0|T|common: [app]human attr(box:  606  141  274  655) --> result:
Male
Age18-60
Direct: Front
Glasses: True
Hat: False
HoldObjectsInFront: False
Bag: No bag
Upper: ShortSleeve UpperStride
Lower: Trousers
Shose: No boots
1727627713833|7f937c50e0|T|common: [nn]detectedInfos: 1
1727627713833|7f937c50e0|T|common: [app]human attr(box:  308  188  207  591) --> result:
Female
Age18-60
Direct: Back
Glasses: False
Hat: False
HoldObjectsInFront: False
Bag: ShoulderBag
Upper: ShortSleeve UpperLogo
Lower: LowerPattern Shorts
Shose: No boots
1727627718787|7fa3285020|T|common: [app]sequential out, retVal: -0x0

e. Model inference results save to image.

./snnf_run.sh -s YoloV8sPose,imageWriter

# ./snnf_run.sh -s YoloV8sPose,imageWriter
1727627887700|7fafe27020|T|common: [app]sequential in
1727627887700|7fafe27020|T|common: [app]warning: sequential model list(not tested)
1727627887725|7fafe27020|I|common: [nn]create model from pluginName: YoloV8sPose takes: 24
1727627887858|7fafe27020|T|common: [app]input image name: resource/image/pose_input.jpg
1727627888707|7f9db2a0e0|I|common: [nn]picked: 5
1727627888707|7f9db2a0e0|I|common: [nn]plot:  0  91%, [(852, 142) - (1169, 753)], person
1727627888710|7f9db2a0e0|I|common: [nn]plot:  0  89%, [(1689, 187) - (1835, 642)], person
1727627888711|7f9db2a0e0|I|common: [nn]plot:  0  89%, [(61, 123) - (232, 601)], person
1727627888711|7f9db2a0e0|I|common: [nn]plot:  0  88%, [(1337, 330) - (1441, 679)], person
1727627888712|7f9db2a0e0|I|common: [nn]plot:  0  87%, [(369, 252) - (480, 671)], person
1727627888839|7f9db2a0e0|T|common: [app]write an image: detected_1883_0931_1727627888712.jpg
1727627893526|7fafe27020|T|common: [app]sequential out, retVal: -0x0

Results will save to the image detected_1883_0931_1727627888712.jpg.

How to build SNNF

Cross-compile for C3V environment.

a. Please use snnf_build.sh for SNNF compiling.

b. All the resource will installed to release folder.

Copy release folder to the C3V platform.
Setup environment variable.

a. Setting environment variables independently.

export LD_LIBRARY_PATH=${PWD}/lib:${PWD}/thirdparty/opencv4/lib:${PWD}/thirdparty/pytorch/lib:${PWD}/thirdparty/freetype/lib:${PWD}/thirdparty/libpng/lib:${LD_LIBRARY_PATH}

b. Run snnf_run.sh will auto set environment variables.

Then, you can run snnf_run.sh for SNNF sample.

Models reference

Model Name	Version or Path
Yolov5s	https://github.com/ultralytics/yolov5/releases/download/v7.0/yolov5s.pt
Human Attributes	https://bj.bcebos.com/v1/paddledet/models/pipeline/PPLCNet_x1_0_person_attribute_945_infer.zip
Light Face	https://github.com/Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB/blob/master/models/onnx/version-slim-320.onnx
Optical character recognition	ch_PP-OCRv3_det_infer ch_ppocr_mobile_v2.0_cls_infer ch_PP-OCRv3_rec_infer
Age recognition	https://gitcode.com/smahesh29/Gender-and-Age-Detection/commits/master 7c024d9d453c9b35a72a984d8821b5832ef17401
Yolov8 Detection	https://github.com/ultralytics/ultralytics
Yolov8 Pose	https://github.com/ultralytics/ultralytics
Yolov8 OBB	https://github.com/ultralytics/ultralytics
Yolov8 Segmentation	https://github.com/ultralytics/ultralytics
Yolov8 Classification	https://github.com/ultralytics/ultralytics
Vehicle attributes	https://bj.bcebos.com/v1/paddledet/models/pipeline/vehicle_attribute_model.zip
License plate recognition	CCPD2020 Yolov8 CCPD detection OCR
Yolov10 Detection	https://github.com/THU-MIG/yolov10
RTMDet	https://github.com/open-mmlab/mmdetection
Face Recognition	https://github.com/deepinsight/insightface
Object tracking	BotSort
Falling Recognition	STGCN

V1.1.0 of SNNF