---
language: en
license: mit
tags:
- vision
- image-segmentation
- sam
pipeline_tag: image-segmentation
datasets:
- custom
metrics:
- accuracy
model-index:
- name: SAM for Image Segmentation
  results:
  - task:
      type: image-segmentation
      name: Image Segmentation
    dataset:
      name: Custom
      type: custom
    metrics:
      - type: accuracy
        value: N/A
        name: Accuracy
---

# SAM (Segment Anything Model) for Image Segmentation

This is a modified version of Facebook's SAM model (sam-vit-base) properly configured for image segmentation tasks in CVAT.

## Model description

SAM can perform segmentation on any object in an image. This model outputs binary segmentation masks.

### Labels

- object: Generic object segmentation