Browse categories
English
$
USD
Connect to your AI
I will build custom yolo object detection and tracking for drones and jetson
Faris
Level 1
About this gig
Getting a vision model to work in a demo is easy. Getting it to run fast and accurately on a real drone, camera or edge device is the hard part, and that is what I build.
I'm an AI and robotics engineer who builds object detection and tracking systems, from a quick model setup to a fully optimized pipeline on Jetson hardware. Example: a YOLO and TensorRT pipeline running at about 200 FPS on a Jetson Orin NX.
What I can build for you:
- Custom object detection: YOLO models trained on your own objects, not just cars and people
- Tracking and analytics: counting, movement tracking, zones and alerts
- Edge optimization: TensorRT and NVIDIA DeepStream so models run fast on Jetson Nano, Xavier NX, Orin, Raspberry Pi with Hailo, or custom drone payloads
- Drone integration: live camera feeds and companion computers on ArduPilot and PX4 builds, and DJI drones with SDK support
You get clean, commented Python or C++ code, the trained model files and a short guide to run everything.
Please message me before ordering with your hardware, camera, target objects and frame-rate goal so I can scope it exactly.
Get to know Faris
Faris
AI Developer, Robotics Engineer, FPV Drone Design Expert
Level 1
- FromPakistan
- Member sinceJun 2026
- Avg. response time1 hour
- Last delivery2 months
Languages
English
I'm an aerospace and AI engineer who designs, simulates and builds drones and robots end to end. I offer UAV and VTOL design in CAD, CFD and FEA analysis, ROS 2 and Gazebo simulation, custom electronics (PCB schematic and layout), and computer vision with YOLO and OpenCV on Jetson or Raspberry Pi. I also consult on flight controllers such as ArduPilot and PX4. You get clear files, honest timelines and work that is checked before delivery. Send me a message with your project and what it needs to do.
My Portfolio
FAQ
What drone platforms are compatible with your AI models?
I work with most platforms supporting companion computers (Jetson/Raspberry Pi). This includes custom ArduPilot/PX4 builds and DJI drones with SDK support. As long as we can access the camera feed or mount a payload, I can integrate the AI.
Do I need to provide the dataset for custom object detection?
For custom objects, providing your own images/videos yields the best results. If you don't have a dataset yet, send me a message! I can often help source images, annotate them, or even generate synthetic data for training.
What hardware do I need to run your optimized models?
For high FPS, I recommend NVIDIA Jetson modules (Nano, Xavier NX, Orin) using TensorRT. However, I can also optimize for Raspberry Pi, Hailo-8L accelerators, or standard PCs depending on your project's power and weight limits.
