# Dual-Mode Beauty Switching Mechanism of LetMagic Beauty SDK in Flutter Apps with ZEGO Audio & Video SDK In modern video social applications, beauty retouching is a core feature for enhancing user experience. Especially for one-on-one real-time video scenarios, developers are challenged to deliver natural, smooth and distortion-free beauty effects, while enabling users to switch freely between AI-powered automatic beautification and manual parameter adjustment. Leveraging the Flutter cross-platform framework and ZEGO Audio & Video SDK, developers can efficiently build one-on-one video applications with dual-mode beauty functions. This article discusses the switching mechanism between AI beauty and manual beauty from the perspectives of product logic and technical implementation. ## Classification and Application Scenarios of Beauty Modes Beauty functions are generally divided into two categories: AI intelligent beauty and manual custom beauty. Driven by algorithms, AI beauty automatically identifies facial features and adjusts parameters for skin smoothing, whitening, eye enlargement and face slimming to achieve one-click beautification, which caters to users who prioritize convenience and natural visuals. Manual beauty allows users to tweak individual parameters based on personal preferences to meet personalized demands. In video social scenarios, the two modes need to support real-time switching, so users can obtain ideal presentation under varying lighting, shooting angles and network conditions. ## Key Points for Switching Logic Design Smooth switching between AI beauty and manual beauty relies heavily on state management and parameter synchronization. First, the SDK maintains two independent sets of beauty parameters: one updated in real time by AI algorithms, and the other configured manually by users. When mode switching is triggered, the UI shall display the current status promptly and ensure seamless transition of beauty effects in video streams. For instance, switching from AI mode to manual mode keeps the existing visual effect without abrupt changes. When switching back from manual mode to AI mode, the system reactivates facial recognition and intelligent optimization to override previous manual settings. ## State Synchronization and User Experience Optimization It is recommended to set AI beauty as the default mode to lower the learning cost for users. Once users adjust any parameter such as skin smoothing intensity or face slimming level, the system automatically switches to manual mode, with other AI-recommended parameters retained as baseline values. Meanwhile, the UI should clearly indicate the active mode via highlighted buttons, text prompts and other cues to prevent confusion. To deliver a more fluent experience, gradual transition is adopted during mode switching for smoother effect changes. ## Performance and Compatibility Considerations When integrating the audio, video and beauty SDKs in Flutter, ensure consistent performance across different platforms. The mode switching process should minimize resource consumption to preserve the clarity and real-time performance of video calls. In addition, conduct comprehensive compatibility tests targeting cameras of various devices and different system versions. This guarantees accurate AI recognition and responsive manual adjustments, delivering stable experience on all device models. ## Conclusion To implement switching between AI beauty and manual beauty in one-on-one video social apps, solid technical support and user-oriented interactive logic are both indispensable. Proper parameter management and state synchronization help deliver intelligent and personalized beauty experiences, thus improving app competitiveness and user retention. With continuous algorithm iteration and hardware upgrades, real-time video beauty will become more refined and scenario-oriented, unlocking more possibilities for the video social industry.