wardrowbe
Self-managed wardrobe management application supports clothing photo management, AI automatic analysis, weather and wear recommendations, family members and wearing records.
Lazy cat OIDC login has been integrated. Wardrobe is a self-hosted wardrobe management application that records photos of clothing, automatically analyzes clothing attributes, generates wearing suggestions, and tracks daily wearing history. The application has been adapted to the Lazy Cat micro service, supporting the Lazy Cat account secret login, data persistence, image uploading, and the Lazy Cat network disk file selector. Suitable for these scenarios: - Organize your clothes into a searchable digital wardrobe. - After uploading clothes photos, AI is used to automatically identify information such as type, color, material, and style. - Generate dressing suggestions based on weather, occasion, and personal preferences. - Record the matching you wear and check the frequency of clothing use and wearing history. Press the guidance to complete the basic settings when entering for the first time: - Create or join a family. - Set the location for weather-related wearing suggestions. - Set common occasions and personal style preferences. - Add the first piece of clothing. ##Configuring AI AI analysis and wearing recommendations require the configuration of OpenAI-compatible API. Enter: `Settings -> AI Endpoints` Fill in an AI Endpoint: - **Name**: Custom name, such as `gpt`. - **URL**: API address, such as `https://api.apimart.ai/v1`. - **API Key**: The key provided by the service provider. - **Vision Model**: A visual model for image recognition. It has been verified that `gpt-4o` can be used for image analysis. - **Text Model**: A model used for text generation and dress recommendations, such as `gpt-4o`. After filling in, proceed in order: 1. Click `Test Connection` to confirm that the API address and Key are available. 2. Click Save to save the configuration. 3. Upload or reanalyze clothing. Notes: - Test Connection mainly verifies whether/models are accessible. It does not mean that Vision Model must support image input. - If image analysis fails, please first check whether Vision Model supports the image message format of `/chat/completions`. - For apimart,`gpt-4o-image` will return `model_not_found`, and `gpt-4o` should be used as the Vision Model. - Currently, API Keys are stored in application preference data and are suitable for self-managed use by individuals or families. ##Add clothing Enter: `Wardrobe -> Add Item` There are two options: - **Single Item**: Add a single item of clothing. - **Bulk Upload**: Upload multiple clothes pictures in batches. Upload source: - **Open from local device**: Select a picture from the current device. - **Open from LazyCat Drive**: Select pictures from LazyCat's online drive. ##Re-analyze clothing If the AI configuration is adjusted and the existing clothing display analysis fails, you can click `Retry` on the clothing card on the Wardrobe page or trigger the analysis again. #Create a dress suggestion Enter: `Suggest Outfit` It is recommended to complete it before use: - Add multiple pieces of clothing. - Set location and weather preferences. - Set Style Profile, Occasions, Weather Sensitivity in Settings. - Configure available AI endpoints.
