Export players and ball from image using any vision API

Project Description: I am looking for a freelancer who can export players and a ball from an image using any vision API, preferably Google Vision, roboflow or amazon rekognition. The src image is JPEG in base64 and the final returned image needs to be a transparent png in base64 so i can apply it to an html canvas. The output inference image needs to be just the players and ball in transparent png format. (I have the opencv script that does this) So, the process needs to be: 1. Send image as base64 var using ajax to 3rd part platform like google vision or roboflow or amazon to be inferenced 2. Receive inferenced image back with players and ball 3. Convert the inferenced image to transparent png that contains just the players and ball (I have a script to do this already using opencv) The process I am using at the moment is with flask and an opencv script however it takes over 20 seconds to inference the image. I need this to be instant, and looking into google vision, amazon or roboflow it looks like i can can the usage of flask and allow the 3rd part platforms to do the work. Skills and Experience: - Proficiency in using vision APIs, specifically Google Vision - Strong understanding of image processing and object detection algorithms - Experience with data extraction and conversion to JSON format - Familiarity with image formats, particularly JPEG - Ability to work with large datasets and handle complex image analysis tasks - Attention to detail and accuracy in extracting and organizing the data