paper summaries
Segment Everything Everywhere All at Once - Summary
The paper presents SEEM, a promptable, interactive model for Segmenting Everything Everywhere all at once in an image. It introduces a versatile prompting engine for different types of prompts, including points, boxes, scribbles, masks, texts, and referred regions of another image. The model can ef