Virtual Visual Object Voice Filtering With Alternative Animations

The patent introduces an intermediary filtering system between the user's input and the avatar's output. The processor acts as a mediator that receives user input, filters out inappropriate expressions through identification and replacement mechanisms, and generates appropriate animations. This intermediary layer allows users to maintain expression freedom while preventing harmful content from reaching other users in the virtual environment.

The patent converts harmful inappropriate expressions into beneficial appropriate expressions through automatic word replacement. When the processor identifies inappropriate words (such as swear words or sexually degrading behavior), it replaces them with appropriate alternative words or phrases. This transformation turns potentially harmful user expressions into safe, acceptable content that maintains communication effectiveness while eliminating discomfort for other users.

The patent segments the complex processing system into distinct functional modules: voice information acquisition module, word identification module, word replacement module, and animation generation module. Each module performs a specific function in the pipeline. This segmentation makes the system more manageable and easier to implement, as each module can be developed, tested, and optimized independently while working together to achieve the overall goal of filtering inappropriate expressions.

The system implements self-service through automatic word identification and replacement without requiring manual intervention. The processor autonomously detects inappropriate expressions, identifies problematic words, replaces them with appropriate alternatives, and generates corresponding animations. This automation reduces the need for human moderators or manual filtering processes, thereby managing system complexity through intelligent self-regulation rather than requiring complex human-in-the-loop mechanisms.