# @beinginvested on Instagram

- **Type:** Video
- **Original URL:** https://www.instagram.com/p/DXfrkOKgjZc
- **Gondola URL:** https://gondola.cc/posts/64518685-beinginvested-instagram
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/c93cd06c2b.jpg
- **Posted:** 2026-04-24T00:57:06.000+00:00
- **Account Owner:** Being Invested | Biz, Tech & Investing (@beinginvested) — https://gondola.cc/beinginvested

## Caption

Comment DETECT below and I’ll DM you the full breakdown + GitHub links.

Someone fine-tuned one AI model to detect and track anything from a drone camera

Trees in a forest. Cars in traffic. Players on a pitch.
Same model. Different data. Different use case every time.
→ RF-DETR — open-source, real-time object detection
→ Fine-tuned on forest drone data — counts every tree automatically
→ Ran on city traffic — tracks every vehicle through intersections
→ Applied to football — detects players, maps formations
→ Paired with OC-SORT tracker for real-time object tracking across frames
No field teams. No manual surveys. No expensive software.
Just a model, a drone, and the right dataset.
The code is free. The datasets are public.
He’s already building demos for wind turbines, solar panels, and power lines next.

——————-

Hey! Here’s everything on the RF-DETR project I talked about:

RF-DETR Model (open source, Apache 2.0)
→ github.com/roboflow/rf-detr

RF-DETR + Trackers (detection + tracking across video frames)
→ github.com/roboflow/trackers
What he’s done with it so far:
→ Forest tree detection — fine-tuned on NeonTreeEvaluation dataset, counts every tree from drone footage
→ City traffic tracking — fine-tuned on VisDrone dataset, 43.6 mAP@50, outperformed YOLO26 by 12 mAP
→ Football analysis — player detection, team clustering, camera calibration
→ Real-time segmentation — 6 model sizes from Nano to 2XLarge
He paired it with OC-SORT tracker so you can track objects across entire videos, not just detect in single frames.

All the code and datasets are public. You can fine-tune it on your own data for any aerial or drone use case.
Built by @ skalskip92 — follow him for more demos. He’s working on wind turbines, solar panels, and power lines next.​​​​​​​​​​​​​​​​

.
.
.
#AI #tech #drone #aerial #traffic

## Stats

- **Views:** 62,936
- **Likes:** 5,857
- **Shares:** 0
- **Comments:** 4,661

## Tags

drone, aerial, ai, traffic, tech

---
Copyright (c) Gondola