Hi 👋

Who is Pedro C. Neto?

I'm a Machine Learning Researcher at INESC TEC in Portugal working on explainable artificial intelligence for computer vision problems, such as face biometrics and digital pathology. I am also finishing my PhD in "Explainable Artificial Intelligence for (Face) Biometrics". In parallel I have led as PI a project on this topic, conducted research on colorectal cancer diagnosis from WSI (with industry partners) and delivered lectures and practical sessions of ML to PhD students at FEUP, Portugal.

You can find my contributions to science by topic:

Biometrics 🧔🏻 🕵🏻‍♂️

1. Causality-inspired Taxonomy for Artificial Intelligence (Paper)
2. Score Calibration
   2.1. PIC-Score: Probabilistic Interpretable Comparison Score (Paper and Code)
3. Bias on Face Recognition
   3.1. Compressed Models Decompress Race Biases: What Quantized Models Forget for Fair Face Recognition (Paper)
4. Masked and Occluded Face Recognition
   4.1. Focusface: Multi-task contrastive learning for masked face recognition (Paper and Code)
   4.2. Beyond masks: On the generalization of masked face recognition models to occluded face recognition (Paper)
   4.3. My eyes are up here: Promoting focus on uncovered regions in masked face recognition (Paper)
   4.4. OCFR 2022: Competition on occluded face recognition from synthetically generated structure-aware occlusions (Paper and Code/Dataset)
5. Morphing Attack Detection
   5.1. OrthoMAD: morphing attack detection through orthogonal identity disentanglement (Paper and Code)
   5.2. Unveiling the Two-Faced Truth: Disentangling Morphed Identities for Face Morphing Detection (Paper and Code)
6. Presentation Attack Detection
   6.1. Myope Models - Are Face Presentation Attack Detection Models Short-Sighted? (Paper)

Digital Pathology

1. Colorectal Cancer WSI dataset and Cervical Cancer WSI dataset
2. Annotation Guidelines For Pathologists and AI Scientists
   2.1. Annotating for artificial intelligence applications in digital pathology: A practical guide for pathologists and researchers (Paper)
3. Detecting and Grading Colorectal Cancer (CRC) from Whole Slide Images with Mixed-Supervision
   3.1. CAD systems for colorectal cancer from WSI are still not ready for clinical acceptance (Paper)
   3.2. iMIL4PATH: A semi-supervised interpretable approach for colorectal whole-slide images (Paper)
   3.3. A CAD System for Colorectal Cancer from WSI: A Clinically Validated Interpretable ML-based Prototype (Paper)
4. Co-developed the first automatic colorectal cancer diagnosis from WSI prototype in Portugal (News, News2 and News3)

Others

1. Push-up/Pull-up/Squats counter for a hackathon (Code)
2. Street Segmentation (Code)