Sharp Stock

English

简体中文

English

简体中文

AI Smart Tagging ​

Tags are key to achieving multi-dimensional asset retrieval. Leveraging AI intelligent recognition, you can rapidly extract core keywords for resources, significantly reducing the manual workload of tag entry.

1. Tagging Logic Description ​

The AI smart tagging feature operates through the following mechanisms:

  • Resource Parsing: Reads the current asset's corresponding disk source file (audio assets call the audio model, while images and other assets call the vision model).
  • Intelligent Extraction: The model automatically outputs a set of semantic keywords.
  • Automated Entry: The system automatically parses the returned keywords and performs operations to create new tags or associate existing tags.

2. Operation Guide ​

  1. Interface Configuration: Complete the configuration for image and audio interfaces in the [Settings] → [AI Model Configuration] panel.
  2. Single Asset Trigger: Expand the "Tags" section in the detail panel. Ensure that only one asset is currently selected (AI tagging is not triggered in multi-select mode).
  3. Execute Tagging: Click the "Auto" button with the blue AI icon next to the tags area.
  4. Result Confirmation: Upon success, new tags will immediately appear on the asset card and in the detail list, and will be synchronized to the global tag pool.

3. Global Tag Pool Integration ​

  • Precise Reuse: The system automatically compares returns with existing tags in the database. If a keyword returned by the AI already exists, the system will perform a direct association to avoid creating redundant duplicate tags.
  • Dynamic Expansion: If brand new keywords are identified, the system will automatically register them as new tags in the library and associate them with the current asset.

4. Core Advantages & Value ​

  • Categorization Consistency: Similar resources analyzed by AI yield highly similar keyword sets, facilitating later global reorganization.
  • Efficiency Multiplier: Especially for newly imported large batches of art assets, AI tagging can provide an excellent initial tagging scheme that you only need to fine-tune.

5. Notes ​

  • Recognition Precision: The tagging effect directly depends on the capabilities of the selected multimodal model. We recommend choosing a model that fits your project's art style.
  • Cost Control: When using cloud-based AI interfaces, please monitor interface call volume and corresponding billing rules.

TIP

After tagging, if you find individual tags that do not meet project expectations, simply click the × next to the tag to remove it. You can also delete unreferenced orphan tags from the library via [Settings] → [Maintenance & Optimization] → [Clean Tags].