Standard Data Management

ModelArts Standard Data Management offers an efficient and convenient framework for managing and labeling data. It supports various data types such as images, text, audio, and video, covering multiple labeling scenarios like image classification, object detection, audio segmentation, and text classification. It is suitable for AI projects in computer vision, natural language processing, audio/video analysis, and more.

The ModelArts Standard Data Management module is currently undergoing a reconstruction and will be upgraded with new capabilities in the era of large models. Stay tuned for updates.

ModelArts Standard Data Management supports multi-dimensional data management capabilities:

• Dataset management: Provides features for creating datasets, previewing data, and managing dataset versions.
• Data labeling: Offers online labeling capabilities for various scenarios like image classification, object detection, audio segmentation, and text triplets. It also provides intelligent image labeling solutions to enhance efficiency. Team labeling features support collaborative efforts by multiple users, as well as the review and acceptance of labeling tasks.
• Data processing: Provides analysis and processing capabilities such as data cleansing, data validation, data augmentation, and data selection.

Figure 1 Data labeling process