$ show CVHub520/X-AnyLabeling
X-AnyLabeling is a cross-platform desktop application (PyQt6) for AI-assisted annotation of images, videos, and 3D point clouds. It serves as a unified interface for a vast model zoo including YOLO (v5-v26), SAM (1/2/3), Grounding DINO, Florence-2, PP-OCR, and many others, enabling auto-labeling for detection, segmentation, pose estimation, OCR, tracking, and more. The application features a plugin-based model architecture, abstracted inference engines (ONNX, TensorRT, PyTorch), and a comprehensive settings system. It also integrates an Ultralytics YOLO training pipeline and a secure in-app update mechanism. While primarily an application, it has a strong hybrid nature due to its deep integration of research code (vendored Grounding DINO, SAHI) and data pipeline utilities (format converters, COCO evaluators). The codebase demonstrates high craftsmanship in its model abstraction layer and inference engine design, though the sheer volume of integrated models introduces a medium risk of technical drift. It is actively maintained with a large community and extensive test coverage.
The LLM identified 5 files worth highlighting for reuse:
Provides a clean, reusable wrapper around ONNX Runtime. Handles session creation, input/output binding, dynamic shape handling, and provider management (CPU/CUDA/TensorRT). This abstraction decouples model inference from the application logic, making it a highly extractable component for any Python project requiring ONNX inference.
Implements standard Kalman filters (KalmanFilterXYAH, KalmanFilterXYWH) for multi-object tracking. Includes methods for initiate, predict, update, and Mahalanobis gating distance. This is a self-contained, mathematically sound implementation that can be directly extracted and used in any tracking pipeline.
Implements a secure update manifest verification system using Ed25519 signatures. Handles canonical JSON generation, signature verification, chunked download with integrity checks, and artifact selection. This is a non-trivial, security-conscious pattern that is completely self-contained and demonstrates high craftsmanship in secure software distribution.
Generates distinct, visually appealing colormaps for segmentation mask visualization. This is a common computer vision utility implemented cleanly and self-containedly. It solves the recurring problem of assigning consistent, distinguishable colors to class labels in annotation tools.
Contains highly optimized Non-Maximum Suppression (NMS) functions for both axis-aligned and rotated bounding boxes. Includes implementations for standard NMS, soft-NMS, and end-to-end model post-processing. These are performance-critical, reusable utilities for any object detection pipeline.