I will train and deploy custom yolo object detection models with opencv
About this Gig
Need a model that finds objects in your images or video, and runs on real hardware? I train and deploy YOLO object detectors with Python, PyTorch and OpenCV.
What I offer:
- Custom YOLO training on your labelled dataset
- Clear evaluation: precision, recall, mAP and example predictions
- OpenCV inference scripts for images, video files or webcams
- Edge deployment on NVIDIA Jetson, or int8 TFLite quantization for microcontrollers
My experience: I'm an MSc Internet of Things and AI student at the University of Bristol. As team captain of a vision-based self-driving car, I collected data, trained the models and deployed YOLOv5 traffic sign and light detection on an NVIDIA Jetson, running alongside a CNN lane-keeping model. The team won a national second prize in the China Robotics and AI Competition. I've also quantized a 1D convolutional autoencoder to int8 for TFLite Micro.
Please message me before ordering with your dataset size, classes and target hardware so I can confirm scope and timeline.
Expertise:
Image processing
•
Object detection
•
Anomaly detection
Programming language:
Python
Tools:
Jupyter Notebook
•
OpenCV
•
TensorFlow
Frameworks:
PyTorch
