Guided SAM: Label-Efficient Part Segmentation

TL;DR

Guided SAM introduces a method that uses learned positional prompts from coarse annotations to efficiently guide the Segment-Anything Model for precise object part segmentation, significantly reducing labeling effort.

Contribution

The paper presents Guided SAM, a novel approach that conditions SAM on learned region prompts from minimal annotations to improve part segmentation accuracy.

Findings

Improves average IoU from 0.37 to 0.49 on car parts dataset.

Reduces annotation effort by a factor of five.

Enhances segmentation efficiency with minimal labeled data.

Abstract

Localizing object parts precisely is essential for tasks such as object recognition and robotic manipulation. Recent part segmentation methods require extensive training data and labor-intensive annotations. Segment-Anything Model (SAM) has demonstrated good performance on a wide range of segmentation problems, but requires (manual) positional prompts to guide it where to segment. Furthermore, since it has been trained on full objects instead of object parts, it is prone to over-segmentation of parts. To address this, we propose a novel approach that guides SAM towards the relevant object parts. Our method learns positional prompts from coarse patch annotations that are easier and cheaper to acquire. We train classifiers on image patches to identify part classes and aggregate patches into regions of interest (ROIs) with positional prompts. SAM is conditioned on these ROIs and prompts. This approach, termed `Guided SAM', enhances efficiency and reduces manual effort, allowing effective part segmentation with minimal labeled data. We demonstrate the efficacy of Guided SAM on a dataset of car parts, improving the average IoU on state of the art models from 0.37 to 0.49 with annotations that are on average five times more efficient to acquire.