ALICE: AI 驅動的 YOLO 數據集全方位管理工具 | ALICE: AI-Powered All-in-One YOLO Dataset Management Toolkit
🔎 工具速覽 / AT A GLANCE
| Category | AI 數據集管理工具 / AI Dataset Management Tool |
| Pricing | 免費開源 (CC BY-NC 4.0) / Free & Open Source |
| BestFor | 需要自定義場景訓練 YOLO 模型的開發者與 Frigate NVR 用戶 / Developers training custom YOLO models and Frigate NVR users |
| GitHub Stars | ⭐ 138 |
🚀 引言 / Introduction
ALICE 是一款專為 YOLO 模型設計的一站式工具包，將分析 (Analyse)、學習 (Learn)、攝取 (Ingest)、策劃 (Curate) 與導出 (Export) 完整整合。它解決了開發者在構建特定場景數據集時，面臨的標記繁瑣與數據冗餘痛點。 | ALICE is an all-in-one toolkit designed for YOLO models, fully integrating Analyse, Learn, Ingest, Curate, and Export. It solves the pain points of tedious annotation and data redundancy faced by developers building scene-specific datasets.
🛠️ 核心功能 / Key Features
提供完整的畫布編輯器，支持標記框的繪製、縮放、移動與刪除，並可按類別或數據集分割（訓練/驗證）進行篩選。 | A full canvas editor for drawing, resizing, moving, and deleting bounding boxes, with filtering by class or dataset split (train/val).
直觀的縮略圖網格視圖，配備數據集統計面板，即時掌握圖像總數、類別分佈及標記覆蓋率。 | Intuitive thumbnail grid view with a statistics panel to monitor total images, class distribution, and annotation coverage in real-time.
基於感知哈希 (pHash) 與 DCT 算法的重複影像檢測，支持多進程加速與側向對比，有效清理冗餘數據。 | Duplicate image detection based on perceptual hashing (pHash) and DCT, featuring multi-processing acceleration and side-by-side comparison to clean redundant data.
與 Frigate NVR 深度集成，可實時瀏覽事件快照，並快速將其轉移至訓練數據集。 | Deep integration with Frigate NVR, allowing real-time browsing of event snapshots and quick transfer to training datasets.
💡 技術亮點 / Tech Highlights
將數據攝取到模型導出的整個生命週期整合在單一工具中，極大提升訓練效率。 | Integrates the entire lifecycle from data ingestion to model export into a single tool, greatly enhancing training efficiency.
透過 pHash 技術與 NMS 邏輯，解決同一相機拍攝的相似影像問題，確保數據集多樣性。 | Solves the issue of similar images from the same camera using pHash and NMS logic, ensuring dataset diversity.
支持一鍵式 builder 腳本及 Docker Compose 部署，快速建立獨立的運行環境。 | Supports one-click builder scripts and Docker Compose deployment for rapid environment setup.
📦 快速上手 / Quick Start
運行 `python3 builder.py` 自動構建虛擬環境並安裝依賴。 | Run `python3 builder.py` to automatically build the virtual environment and install dependencies.
執行 `./alice.py` 並訪問 `http://localhost:8080` 啟動管理介面。 | Execute `./alice.py` and visit `http://localhost:8080` to launch the management interface.
在設置中下載 YOLO 模型並配置 Frigate 路徑即可開始工作。 | Download a YOLO model in Settings and configure Frigate paths to get started.
準備好試試 ALICE: AI 驅動的 YOLO 數據集全方位管理工具 | ALICE: AI-Powered All-in-One YOLO Dataset Management Toolkit 了嗎？
Ready to try ALICE: AI 驅動的 YOLO 數據集全方位管理工具 | ALICE: AI-Powered All-in-One YOLO Dataset Management Toolkit?
前往 GitHub 頁面 →
留言
張貼留言