ImageLabel is a macOS image annotation tool for machine learning, computer vision, and dataset creation. It integrates project management, fine-grained annotation, model pre-annotation, and multi-format export into a clean and efficient workspace, helping you prepare object detection and instance segmentation datasets more easily. 【Flexible Annotation Tools】 Supports rectangular and polygonal borders and custom graphics for more accurate annotations. 【Efficient Image and Progress Management】 Quickly search and filter images by filename, tag, or annotation status. Mark images as completed, pending review, or unprocessed. Project statistics visually display completion progress, number of annotations, type distribution, and label distribution, allowing you to easily monitor the dataset status. 【Local Core ML Pre-annotation】 Import your own Core ML object detection model for offline recognition, generating bounding boxes for the current image or the entire dataset in batches. Supported model formats: mlmodel, mlpackage, mlmodelc 【Compatible with common training workflows】 Annotation results can be exported as: Create ML JSON, COCO JSON, LabelMe JSON. Whether you use Create ML, YOLO, Detectron2, Mask R-CNN, SAM, or other compatible tools, you can smoothly connect to subsequent training workflows. 【More Useful Features】 - Image cropping and saving copies without overwriting the original - Customize annotation lines, fill effects, and label styles - Support common image formats such as JPG, PNG, HEIC, TIFF, and BMP - Separate management of projects and image materials - Quick access to recent projects - Local saving of annotation and processing progress - User Agreement https://my.feishu.cn/docx/Gje7dO42IoNrMJxEK8KctgoRnvb
Developer Tools · iPhone, iPad · Free
