Héctor Martínez Cabezas

https://theroboticsclub.github.io/2026-internship-Hector_Martinez/Héctor Martínez CabezasA minimal, responsive and feature-rich Jekyll theme for technical writing. 2026-10-07T14:04:50+02:00 Héctor Martínez Cabezas https://theroboticsclub.github.io/2026-internship-Hector_Martinez/ Jekyll © 2026 Héctor Martínez Cabezas /2026-internship-Hector_Martinez/assets/img/favicons/favicon.ico /2026-internship-Hector_Martinez/assets/img/favicons/favicon-96x96.png Testing the End-to-End Visual Control Model and Fine-Tuning MobileNetV1-SSD with ONNX Export2026-10-07T00:00:00+02:00 2026-10-07T11:56:10+02:00 https://theroboticsclub.github.io/2026-internship-Hector_Martinez/posts/post_1/ Hector

This week, I finished working on the End-to-End Visual Control exercise by testing the trained model. I also continued improving the fine-tuning of the pre-trained MobileNetV1-SSD model and created a script to export the newly trained model to ONNX for further testing. Testing End-to-End Visual Control After exporting the onnx model last week, I tested it in Unibotics by following the instruc...

Training a CNN Model for End to End Visual Control and Fine-Tuning Visual Human Detection2026-09-30T00:00:00+02:00 2026-09-30T00:00:00+02:00 https://theroboticsclub.github.io/2026-internship-Hector_Martinez/posts/post_0/ Hector

For my first two internship tasks, I had to work on two exercises in Unibotics: End to End Visual Control and Visual Object Detection. End to End Visual Control. Creating the CNN and training the model In this exercise, a car must complete a lap by following a red line. The exercise documentation requires using an AI model trained with a CNN (Convolutional Neural Network). Because of this, I...