K230 AI Demo introduction

K230 AI Demo

Overview

K230 AI Demo integrates face, human body, hand, license plate, word continuation and other modules, including classification, detection, segmentation, recognition, tracking and other functions, to provide customers with a reference on how to use K230 to develop AI-related applications.

Hardware environment

• K230-UNSIP-LP3-EVB-V1.0/K230-UNSIP-LP3-EVB-V1.1/CanMV-K230 (supported by default)

Source code location

The source code path is located insrc/reference/ai_poc and the directory structure is as follows:

# Detailed Demo documentation is available in the AI Demo subdirectories (eg: bytetrack, face _ detection, etc.)
.
├── build_app.sh
├── bytetrack
├── cmake
├── CMakeLists.txt
├── face_detection
├── face_emotion
├── face_gender
├── face_glasses
├── face_landmark
├── face_mask
├── face_parse
├── face_pose
├── falldown_detect
├── finger_guessing
├── fitness
├── head_detection
├── k230_bin
├── licence_det
├── licence_det_rec
├── llamac
├── nanotracker
├── object_detect_yolov8n
├── person_attr
├── person_detect
├── pose_detect
├── pphumanseg
├── segment_yolov8n
├── shell
├── smoke_detect
├── space_resize
├── sq_hand_det
├── sq_handkp_class
├── sq_handkp_det
├── sq_handkp_ocr
├── sq_handreco
├── traffic_light_detect
├── vehicle_attr
└── version

The kmodel, image, and related dependency paths are located in src/big/kmodel/ai_poc. The directory structure is as follows:

.
├── images
│   ├── 1000.jpg
│   ├── 1024x1331.jpg
│   ├── 1024x624.jpg
│   ├── 1024x768.jpg
│   ├── 640x340.jpg
│   ├── bus.jpg
│   ├── bytetrack_data
│   ├── car.jpg
│   ├── falldown_elder.jpg
│   ├── hrnet_demo.jpg
│   ├── input_hd.jpg
│   ├── input_ocr.jpg
│   ├── licence.jpg
│   ├── smoke1.jpg
│   └── traffic.jpg
├── kmodel
│   ├── bytetrack_yolov5n.kmodel
│   ├── cropped_test127.kmodel
│   ├── face_detection_320.kmodel
│   ├── face_detection_640.kmodel
│   ├── face_emotion.kmodel
│   ├── face_gender.kmodel
│   ├── face_glasses.kmodel
│   ├── face_landmark.kmodel
│   ├── face_mask.kmodel
│   ├── face_parse.kmodel
│   ├── face_pose.kmodel
│   ├── hand_det.kmodel
│   ├── handkp_det.kmodel
│   ├── hand_reco.kmodel
│   ├── head_detection.kmodel
│   ├── human_seg_2023mar.kmodel
│   ├── licence_reco.kmodel
│   ├── LPD_640.kmodel
│   ├── nanotrack_backbone_sim.kmodel
│   ├── nanotracker_head_calib_k230.kmodel
│   ├── ocr_det.kmodel
│   ├── ocr_rec.kmodel
│   ├── person_attr_yolov5n.kmodel
│   ├── person_detect_yolov5n.kmodel
│   ├── person_pulc.kmodel
│   ├── traffic_detect_yolov5s_best.kmodel
│   ├── vehicle_attr_yolov5n.kmodel
│   ├── vehicle.kmodel
│   ├── yolov5n-falldown.kmodel
│   ├── yolov5s_smoke_best.kmodel
│   ├── yolov8n_320.kmodel
│   ├── yolov8n_640.kmodel
│   ├── yolov8n-pose.kmodel
│   ├── yolov8n_seg_320.kmodel
│   └── yolov8n_seg_640.kmodel
└── utils
    ├── Asci0816.zf
    ├── bu.bin
    ├── dict_6625.txt
    ├── HZKf2424.hz
    ├── jiandao.bin
    ├── llama.bin
    ├── shitou.bin
    └── tokenizer.bin

Compile and run the program

• Enter the src/reference/ai_poc

• Execute the build_app.sh script (before executing the script, make sure that there are corresponding kmodels, images, and utils under the src/big/kmodel/ai_poc), this will generate a unified copy of kmodel, images, utils, shell, and elf to the k230_bin folder

• Copy the entire folder to the board, and execute the sh script on the big core to run the corresponding AI demo