Working Experience

Evocortex GmbH: Machine Learning [2022-2024]

The following activities were made at Evocortex robotics company on the Machine Learning department in Nürnberg, Germany between 2022-2024. The Bachelor thesis is described on the Publications section.

To notice: I trained object detection models and then wrote my thesis on the first entire year, so the activities of the second year are shown next.

Problem: using the bachelor thesis’s discoveries, the synthetic data rendered for training object detection models were still not strong enough to make the detection model to work on real life at an acceptable industrial requirement.

• Task 1: internal structure analysis of YOLO-based based neural networks to implement saliency maps that could explain the network’s focus point on the image during the detection task, implementing both a gradient-based-method algorithm (Gradcam based) and a non invasive one (D-RISE).
• Task 2: pipeline implementation for synthetic data generation using diffuse models based on Stable Diffusion v1.5 and ControlNet neural network (NN) structure.
  • Usage of the Low Rank Adaptation technique for training SD models, prompt experimentation and testing by training object detection models.
  • The usage of Generative Adversarial Networks (GANs) was also followed, but the result obtained was not strong enough.

Omega: Improvement department [2024-2025]

The following activities were made at Omega company on the Improvement department in Monterrey, Mexico between 2024-2025

Problem: Omega is a local manufacturer of electric heaters and their online presence was not strong enough to compete with other businesses that are implementing digital business strategies.

• Task 1: Web page made from scratch. New Omega Webpage
• Task 2: Social media and automating quotation process.
• Task 3: responsible of machinery purchase, improvements and small automation tasks using PLCs.