Labelme

Labelme is a Python-based image annotation tool used to create computer vision datasets. It serves as a visual editor for semantic segmentation, allowing users to define object boundaries using polygons, rectangles, points, and circles. The application also functions as a multispectral image annotator, supporting high-bit depth TIFF files used in satellite and scientific imagery.

The tool incorporates AI-assisted labeling capabilities to automate the creation of masks and polygons. These features allow for shape generation driven by text prompts or interactive point selections, which propose boundaries based on user-placed positive and negative points.

The software covers a broad range of data management and annotation tasks, including the creation of dense pixel masks, rotated bounding boxes, and video frame sequencing. It includes a pipeline for translating internal JSON state persistence into standard dataset formats such as COCO and Pascal VOC. Additional capabilities include image-level classification flags, geometry refinement tools, and batch image importing.

Features

AI-Assisted Labeling - Implements an annotation interface that uses AI models to generate masks and polygons from text prompts or point selections.

Computer Vision Annotation - Provides a visual editor for creating high-quality annotated datasets of images and videos for vision AI.

AI Model Integrations - Integrates external machine learning models into the UI to generate polygons and masks from text or points.

Semantic Segmentation - Provides a visual editor for creating precise pixel-level masks and polygonal boundaries for image segmentation tasks.

JSON-Based Persistence - Persists image metadata, shape coordinates, and labels as structured JSON files on the local filesystem.

Vision Dataset Converters - Provides a pipeline for translating internal JSON annotation data into standard COCO and Pascal VOC formats.

Image Annotation Tools - Provides a desktop application for labeling images and videos with polygons, rectangles, and points.

Vision Dataset Standardizers - Converts custom annotations into standardized vision formats like VOC or COCO for consistent model training.

Visual Annotation Tools - Renders saved JSON annotation data as visual overlays to verify label accuracy.

Data Format Translators - Converts internal JSON annotation representations into standard computer vision dataset formats like COCO and Pascal VOC.

Interactive Shape Proposals - Implements an interactive vision model that proposes shape boundaries based on user-placed positive and negative points.

Text-Driven Shape Detection - Detects and proposes shapes for every instance of a specific class in an image using a text-driven detector.

Multispectral Annotators - Functions as a specialized annotator for high-bit depth and multispectral TIFF files used in satellite imagery.

Multispectral Processors - Supports the loading and annotation of high-bit depth or multispectral TIFF files used in satellite imagery.

Semantic Video Segmentations - Labels sequences of images within a video to create datasets for semantic segmentation.

Raster Mask Editors - Creates raster-based masks combined with bounding boxes for precise pixel-level segmentation.

Annotation - Polygonal annotation tool for images.

Computer Vision Tools - Versatile annotation tool for image segmentation and labeling.

Image Annotation - Python-based polygonal annotation for various geometric shapes.

Image Annotation Tools - General purpose image annotation tool supporting various formats.

Data Annotation Tools - Polygon annotation tool for creating image segmentation datasets.

wkentarolabelme

Features

Star history