About Pilar
Spanish
Native or bilingual
English
Fluent
German
Conversational
Experience
- Panacea Cooperative ResearchArtificial Intelligence ResearcherNovember 2022 - December 2024 (2 years and 1 month)Granada, Spain• Principal researcher in a CBCT‑to‑CT image translation project using deep neural networks such as Pix2Pix, CycleGAN, and CUT, with the goal of generating CT images realistic enough to allow accurate segmentation of bone and facial structures. Results included segmentations with a mean Dice coefficient of 0.915.• Automatic classification of simulated skull‑face overlays into correct or incorrect using various classic deep classification networks, achieving an accu‑ racy of 97.8%.• Automatic analysis of facial features in simulated images to improve the robustness of an automatic skull‑face overlay algorithm. This work optimized reliability under more varied and realistic conditions.• Principal developer of a system to estimate camera‑to‑subject distance using facial indices and models such as Random Forest and Gradient Boost‑ ing, reducing the mean absolute error from 32 cm (state of the art) to 5.6 cm.• Designed a system for automatic exclusion of negative candidates based on facial index comparison and prediction intervals estimated using various machine learning algorithms. Achieved reliable exclusion of over 60% of negative cases under more variable conditions than those tested in the previous study.• Provided guidance and supervision to an intern student.• Presented research results at national and international conferences for non‑technical audiences.
- Wildlife.aiData ScientistNovember 2022 - May 2024 (1 year and 6 months)New Zealand• Trained YOLO models to detect and classify fish species in underwater footage. Performed data quality and model performance analysis. Produced statistical summaries for biologists.• Provided support in the development of interactive Jupyter Notebooks to simplify code usage for non‑technical biologists.
- FruitpunchAIAI engineerApril 2023 - July 2023 (3 months)• AI for European wildlife challenge: As part of the "model exploration" team, conducted research and applied various deep learning models (YOLOv8, YOLO‑NAS, and multiscale vision transformers) to detect and classify European wildlife in camera trap images.• Contributed to the publication of project results.
Recommendations
Be the first to recommend Pilar
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Bachelor's Degree in Mathematics and Computer ScienceUniversity of Granada2022Bachelor's Degree in
- Bachelor's Degree in Computer ScienceUniversity of Duisburg‑Essen