Kategorien durchsuchen
Entdecken
Fiverr Pro
Deutsch
$
USD
I will develop edge ai and tinyml on esp32 and raspberry pi
IoT Consultant and Fullstack Developer
"IoT Consultant & Fullstack Developer bridging the gap between hardware and the web. Backed by Computer Engineering at UFRJ and high-stakes competitive robotics, I build systems designed to perform un...
Über diesen Service
Bring Your Hardware to Life with Edge AI
Cloud AI is too slow, expensive, and internet-dependent for real-time physical products. If you are building the next generation of IoT devices, you need intelligence on the edge.
I specialize in deploying Machine Learning (TinyML) and Computer Vision models directly onto ESP32 and Raspberry Pi.
My code is forged in the high-stakes environment of competitive robotics. When hardware interacts with the physical world, a millisecond delay or a memory leak means failure. I bring this exact engineering rigorbacked by solid C/C++ architectures and native Linux developmentto your product. I don't just train a model; I ensure it runs flawlessly on memory-constrained microcontrollers.
What I Can Build For You:
- Sensor AI (ESP32): Predictive maintenance, anomaly detection, and gesture recognition using IMU/accelerometer data.
- Audio AI (ESP32): Keyword spotting and audio classification without cloud APIs.
- Vision AI (Raspberry Pi): Real-time object detection and computer vision using Python and OpenCV.
- End-to-End Pipeline: Data collection guidance, model training (Edge Impulse/TF Lite), and optimized C++ deployment.
Let's make your hardware smart, offlin
Plattform:
ESP32
Expertise:
Firmware-Entwicklung
•
Bildverarbeitung
•
AI
Mein Portfolio
FAQ
Do I need to ship my custom hardware to you?
No. I develop and test the logic using standard development boards (ESP32 DevKit, Raspberry Pi) on my workbench. Once the system is validated, I deliver the firmware/scripts and guide you on how to flash it onto your custom board remotely.
Can the ESP32 handle facial recognition or heavy video processing?
No, the ESP32 lacks the RAM for complex video processing. It is perfect for sensor data (IMU, temperature) or basic audio keyword spotting. For Computer Vision, object detection, or facial recognition, please select the Raspberry Pi (Premium) package.
What is the difference between Edge AI and Cloud AI?
Cloud AI relies on a constant internet connection and external servers, which introduces latency. Edge AI (TinyML) runs the neural network locally on the microcontroller's hardware. It operates entirely offline, responds in milliseconds, saves battery, and guarantees data privacy.
How do we get the dataset to train the AI?
If you already have a dataset (CSV, WAV, or images), you can send it to me. If not, I will provide you with a data-forwarding script so you can easily record the physical data from your own sensors directly into Edge Impulse for training.
Will I receive the source code and IP rights?
Absolutely. Every package includes the delivery of the complete, well-documented source code (C/C++ or Python), the trained model library, and full ownership of the intellectual property for your commercial use.
