manuel veras

As a Master's student at UFRGS, I am dedicated to advancing the field of computer vision, particularly in the area of 360-degree object detection. My academic journey has equipped me with a robust understanding of deep learning techniques, which I have effectively applied in the agritech sector to tackle real-world challenges, such as weed detection.

In my current role as a Software Engineer specializing in Machine Learning at Accore Systems Engineering, I have successfully spearheaded the training of a real-time deep learning network for weed detection, achieving over a 2x improvement in speed compared to standard detection networks without compromising performance. My work involves developing algorithms for detecting green vegetation on brown soil using aerial imagery, optimizing network performance through advanced techniques, and designing business-specific performance metrics to evaluate machine learning models.