Dynamic AI Photo Booth Enhancement

I'm looking for an experienced developer to create an innovative AI server tailored specifically for my photo booth business. This server must be capable of delivering real-time image processing with several advanced features. **Key Features Required:** - **Face Detection:** The AI should promptly identify faces within the frame. - **Facial Recognition:** Ability to recognize and differentiate between faces. - **Background Removal:** Instantly remove or replace image backgrounds. - **Face Swap:** Exchange faces between the source photo and an individual taking the photo. - **Theme-Based Modifications:** Apply themed effects based on custom prompts. **Usage scenarios:** The system should allow to process the images in the following options: - face swap: the guests face is transposed onto the body of a source image (eg. my face is placed on a source image of Superman) - the face has to be as close to the original (guest's face) to avoid confusion like "doesn't look like me"; - face swap with custom prompt: similar as above but instead of a ready source image, the processing server takes some input from the user - so for example the photo booth app takes a photo of me, then I write on the screen eg. "superman lifting a truck" and the result is an AI generated image based on this prompt with my face on it [the user input can be passed on from an edit field from the photo booth app, the prompt needs to also have some part that can be input before by the admin and is fixed, while the user input becomes the variable]; - filter/style: image is just changed like using a filter (eg. [login to view URL]), while there should be ways to set new kinds of "themes" based on, most probably, custom prompts. **Requirements** - any face swap features need to have the guest's face on the end image as close to their own as possible (not changed too much to avoid confusion ""doesn't look like me"); - full processing needs to be as fast as possible; - server should run on my own hosting server (provided for development too); - system should have an admin GUI where I can configure new options for each effect, insert (and test) new prompts; - if only possible, system should run on self hosted AI algorithms, not external API based services, in order to cut down processing times; **On event process description:** The photo taking is managed by a photo booth app. After a photo of the guest is taken, it is sent to an external service for post processing, the processing server then returns the processed photo back to the photo booth app, which applies the image template and displays to the user. Postprocessing server is defined via a dedicated link and password combination. **Real-Time Processing:** The solution must handle all image enhancements live during the event to ensure a seamless and engaging user experience. This of course requires fast Internet connection, but the processing server needs to handle the task as fast as possible, so total process (server part) cannot take longer than ~10s. Of course this can depend on the server power, so please propose recommended server service and requirements. **Ideal Skills & Experience:** - Proficiency in AI and Machine Learning, particularly in image processing. - Strong background in developing real-time processing systems. - Familiarity with creating user-friendly servers for event-based applications. I need a solution that isn't just functional but also fast, reliable and can contribute to the overall entertainment of the photo booth experience. The capacity to innovate and troubleshoot on-the-fly will be essential for success in this project. **Examples** [login to view URL] [login to view URL] [login to view URL] [login to view URL] [login to view URL] (this is the closest example of a full system, creating an account is required) **Offer requirements** Simple - just read the full project description. I will read ONLY offers from freelancers that have actually read the description and will skip automated responses. To make sure you have read the description, just start your offer with the word READ. If you have any questions that need clarifying, drop a chat note.