# Segment anything with FastSAM using Ultralytics 🔥

**URL:** https://community.ultralytics.com/t/segment-anything-with-fastsam-using-ultralytics/118
**Category:** Resources
**Tags:** ultralytics-official, news
**Created:** [July 19, 2024, 8:13pm UTC](https://community.ultralytics.com/t/segment-anything-with-fastsam-using-ultralytics/118 "2024-07-19T20:13:51Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![muhammadrizwanm](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/muhammadrizwanm/32/71_2.png) [@muhammadrizwanm](https://community.ultralytics.com/u/muhammadrizwanm)
#### Post date: [July 19, 2024, 8:13pm UTC](https://community.ultralytics.com/t/segment-anything-with-fastsam-using-ultralytics/118/1 "2024-07-19T20:13:51Z")

</div>

💭 Here, I have employed a combination of instance segmentation techniques and the FastSAM model. This real-time, CNN-based solution segments any object in an image using various user prompts. :robot_dance:

**Instance Segmentation with FastSAM using Ultralytics**

| Sample Image 01 | Sample Image 02 |
| --- | --- |
| ![Sample#01](https://us1.discourse-cdn.com/flex001/uploads/ultralytics1/original/1X/5ce330b0afc2e30ce16bda21a802521fc14478a4.jpeg) | ![Sample#02](https://us1.discourse-cdn.com/flex001/uploads/ultralytics1/original/1X/f008a15dd881cb6b06d892fd730ab2dcfc60f533.jpeg) |

Learn more ➡ [FastSAM (Fast Segment Anything Model) - Ultralytics YOLO Docs](https://docs.ultralytics.com/models/fast-sam/)

---

<div class="post-metadata">

### Author: ![glenn-jocher](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/glenn-jocher/32/83_2.png) [@glenn-jocher](https://community.ultralytics.com/u/glenn-jocher)
#### Post date: [July 20, 2024, 3:47pm UTC](https://community.ultralytics.com/t/segment-anything-with-fastsam-using-ultralytics/118/2 "2024-07-20T15:47:58Z")

</div>

So prompt engineering for SAM?

---

<div class="post-metadata">

### Author: ![muhammadrizwanm](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/muhammadrizwanm/32/71_2.png) [@muhammadrizwanm](https://community.ultralytics.com/u/muhammadrizwanm)
#### Post date: [July 22, 2024, 7:16am UTC](https://community.ultralytics.com/t/segment-anything-with-fastsam-using-ultralytics/118/3 "2024-07-22T07:16:52Z")

</div>

@glenn-jocher Absolutely, prompt engineering is crucial for guiding FastSAM in instance segmentation. Here, I’ve combined image segmentation with object tracking for improved results.