The Trainable Weka Segmentation (TWS) tool is a machine learning-based image segmentation tool that allows users to train a classifier using a limited number of manual annotations and then automatically segment the remaining data. It works interactively, allowing users to guide the training by providing corrections to the classifier output. TWS can also perform unsupervised segmentation using clustering and can be customized to use user-designed feature maps or classifiers. The tool is integrated with Fiji and WEKA, providing a user-friendly interface for loading 2D or 3D images and performing automatic segmentation through interactive learning. TWS supports a wide range of image modalities and has been used in various imaging pipelines for tasks such as analyzing wing photomicrographs, visualizing myocardial blood flow, monitoring bee nests, and cell tracking. The tool includes a graphical user interface with training and options panels, allowing users to train classifiers, apply them to images, load/save classifiers and data, and adjust settings. It also supports scripting for automation and provides a macro language compatible with ImageJ. The tool includes a variety of image features for edge detection, texture description, noise removal, and membrane detection, which can be selected and adjusted to suit specific segmentation tasks. The user is encouraged to carefully select the type and size of image features based on the specific application and to provide sufficient training samples to ensure accurate segmentation.The Trainable Weka Segmentation (TWS) tool is a machine learning-based image segmentation tool that allows users to train a classifier using a limited number of manual annotations and then automatically segment the remaining data. It works interactively, allowing users to guide the training by providing corrections to the classifier output. TWS can also perform unsupervised segmentation using clustering and can be customized to use user-designed feature maps or classifiers. The tool is integrated with Fiji and WEKA, providing a user-friendly interface for loading 2D or 3D images and performing automatic segmentation through interactive learning. TWS supports a wide range of image modalities and has been used in various imaging pipelines for tasks such as analyzing wing photomicrographs, visualizing myocardial blood flow, monitoring bee nests, and cell tracking. The tool includes a graphical user interface with training and options panels, allowing users to train classifiers, apply them to images, load/save classifiers and data, and adjust settings. It also supports scripting for automation and provides a macro language compatible with ImageJ. The tool includes a variety of image features for edge detection, texture description, noise removal, and membrane detection, which can be selected and adjusted to suit specific segmentation tasks. The user is encouraged to carefully select the type and size of image features based on the specific application and to provide sufficient training samples to ensure accurate segmentation.