Skip to main content

Python utilities for archive, config and zero-shot detection

Project description

klygo

Thư viện Python hỗ trợ xử lý ZIP, quản lý dataset YOLO, đọc/ghi cấu hình và ảnh, nhận diện vật thể zero-shot, crop và trực quan hóa kết quả.

Cài đặt

git clone https://github.com/IchigoMazone/klygo.git
cd klygo
uv sync

Các package

  • klygo.archive: nén, giải nén, tìm kiếm, kiểm tra, chỉnh sửa, gộp và chia ZIP.
  • klygo.datasets: partition, repartition, unpartition, merge, split và remap dataset YOLO.
  • klygo.io: đọc/ghi YAML, JSON, TOML; đọc ảnh bằng PIL hoặc OpenCV.
  • klygo.models: nhận diện zero-shot trên ảnh, thư mục ảnh và video.
  • klygo.visualize: hiển thị, vẽ bbox, crop, đọc crop và thống kê dataset.

Sử dụng nhanh

Archive

import klygo.archive as ar

ar.compress("dataset", "dataset.zip", overwrite=True)
ar.extract("dataset.zip", output_dir="extracted", overwrite=True)
files = ar.search("dataset.zip", "images/*.jpg")

Dataset YOLO

import klygo.datasets as ds

ds.partition(
    source="raw_dataset.zip",
    target="dataset",
    ratios=(0.8, 0.1, 0.1),
    overwrite=True,
)

info = ds.get_dataset_info("dataset")
print(info)

Đọc cấu hình và ảnh

import klygo.io as io

config = io.Config("config.yaml").read()
pil_images = io.read_images("dataset/images", backend="pil")
opencv_images = io.read_images("dataset/images", backend="opencv")

Model

from klygo.models import Model

model = Model()

predictions = model.predict(
    source="dataset/images",
    prompt="car. person.",
)

detected_frames = model.detect(
    source="traffic.mp4",
    output_path="detected.mp4",
    annotated_dir="annotated",
    dataset_dir="generated_dataset",
    prompt="car. person.",
)

detection_results = model.detect(
    source="dataset/images",
    prompt="car. person.",
    metadata=True,
)

crop_results = model.crop(
    source="traffic.mp4",
    target="crops",
    prompt="car. person.",
)

Visualize

import klygo.visualize as vis

grid = vis.read_crops("crops")  # lưới 5 x 5 mặc định
vis.show_image(grid, title="Detected objects")

saved_detections = vis.read_detections("annotated")

Quy ước đường dẫn

  • source: file hoặc thư mục đầu vào.
  • *_path: đường dẫn file.
  • *_dir: đường dẫn thư mục.
  • target: file hoặc thư mục đầu ra linh hoạt.

Tài liệu chi tiết

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

klygo-2.0.10.tar.gz (3.2 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

klygo-2.0.10-cp312-cp312-win_amd64.whl (1.8 MB view details)

Uploaded CPython 3.12Windows x86-64

File details

Details for the file klygo-2.0.10.tar.gz.

File metadata

  • Download URL: klygo-2.0.10.tar.gz
  • Upload date:
  • Size: 3.2 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.0

File hashes

Hashes for klygo-2.0.10.tar.gz
Algorithm Hash digest
SHA256 60f34e2bc27089659fb7057d986f58852ba082be1b272c5e736f9ac28bef24ba
MD5 8fd0203cbe26316617624627915c3599
BLAKE2b-256 14bb39d9bce3576e5e229504f6f1ef8f496e21e58f638c83ed9fa8c48aaec608

See more details on using hashes here.

File details

Details for the file klygo-2.0.10-cp312-cp312-win_amd64.whl.

File metadata

  • Download URL: klygo-2.0.10-cp312-cp312-win_amd64.whl
  • Upload date:
  • Size: 1.8 MB
  • Tags: CPython 3.12, Windows x86-64
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.0

File hashes

Hashes for klygo-2.0.10-cp312-cp312-win_amd64.whl
Algorithm Hash digest
SHA256 19b2dd32ff9f6cab799e6ec403abdfd6dbf422b4441da11633f920efec92dd78
MD5 13e9141b7361961ef2f4e99c9e825415
BLAKE2b-256 04fe5c205784a867923dfac05d7650342b8fe22ceaad9ddc354b62535dc8d477

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page