- Step1: Install SegAgent via pip: pip install segagent
- Step2: Import and initialize the agent with your OpenAI key and SAM model back end
- Step3: Load an image using SegAgent’s reader utility
- Step4: Send a text prompt to the agent: agent.segment(image, "segment the main object")
- Step5: Review and refine generated masks through follow-up prompts
- Step6: Export final masks in COCO or PNG format