ADVIS-SR: Real-Time Anomaly Detection for Safe Human–Robot Interaction in real industrial scenario EU Project 🇪🇺
Industrial Safety Detector using deep Generative models based Anomaly Detection on real world
Projects & Research Work
This page highlights selected research projects, industrial systems, ongoing collaborations, open-source contributions, and applied machine learning work spanning Explainable AI, Computer Vision, Anomaly Detection, Trustworthy Machine Learning, and Industrial AI .
ADVIS — Industrial AI & Robotics Safety
Research and deployed systems developed around visual anomaly detection, human–robot collaboration, and safety monitoring for industrial environments.
ADVIS-UniGra: RGB Anomaly Detection for Safe Human–Robot Collaboration EU Projects 🇪🇺
RGB-based anomaly detection application for safety monitoring in collaborative robotics environments using synthetic industrial data.
Completed & Featured Research
Selected completed research projects, published methods, and open-source research contributions.
ShapBPT: Image Feature Attributions using Data-Aware Binary Partition Trees Research/XAI
A data-aware XAI method for image feature attribution using Binary Partition Trees and hierarchical Shapley values.
AI on Edge Devices Research/Edge AI
Lightweight AI deployment on edge devices, including Raspberry Pi-based computer vision systems.
Explainable Anomaly Detection Research/XAI, AD
A case study on building trust in anomaly detection systems using VAE-GAN models and explainable AI.
LIME Stratified Sampling Research/XAI
An improved LIME sampling strategy for generating more stable and reliable image explanations.
Freelance & Applied Projects
Selected client-oriented and independent projects applying machine learning, computer vision, image processing, and data analysis to practical problems across different application domains.
Applied and Freelance Research Projects Applied AI
Applied research and freelance projects in medical imaging, computer vision, machine learning, and intelligent decision systems.
Ongoing Research & Collaborations
Active research directions, manuscripts in preparation, collaborative projects, datasets, and open-source research currently under development.
From Reconstruction to Reasoning: The Evolution of Deep Learning Architectures for Video Anomaly Detection Video Anomaly Detection / Deep Learning
A comprehensive study of the evolution of video anomaly detection, from reconstruction and prediction methods to transformers, foundation models, and reasoning-oriented approaches.
Deep Learning for Time-Series Anomaly Detection Time-Series / Anomaly Detection
Research on deep learning and one-class approaches for time-series anomaly detection, including robust training, threshold calibration, and evaluation on benchmark datasets.
Segmentation-Aware Explainable AI with SAM and Shapley Attribution Explainable AI / Computer Vision
Investigating SAM-based hierarchical image segmentation and Shapley attribution for spatially coherent and quantitatively evaluated computer vision explanations.
Quantum Machine Learning for Anomaly Detection Quantum Machine Learning / Anomaly Detection
Exploratory research on variational quantum models and quantum-enhanced representations for anomaly detection.
Explainable Multimodal Visual Question Answering Multimodal AI / Explainable AI
Investigating explainability and faithfulness in multimodal visual question answering using vision-language models and gradient-based attribution methods.
Explainable Deep Learning for Medical Image Analysis Medical Imaging / Explainable AI
Research on explainable deep learning for medical image classification and analysis, combining modern CNN architectures with interpretable prediction methods.
Interpretable Hierarchical Machine Learning for Small-Sample Prediction of Surface and Tribological Responses in Electrical Discharge Machining Interpretable ML / Smart Manufacturing
Interpretable hierarchical machine learning for small-sample prediction and analysis of surface and tribological responses in electrical discharge machining.
EdgeXAD: Visual Anomaly Detection on Edge Devices Edge AI / Visual Anomaly Detection
Developing efficient normal-only visual anomaly detection pipelines for deployment on resource-constrained edge devices and RGB camera systems.
DistriMuSe Industrial Robotics Safety Dataset Industrial AI / Dataset
A multi-view RGB dataset for unexpected-condition detection and human safety in industrial robotic palletizing workflows, including normal and anomalous operating scenarios.
Trustworthy Anomaly Detection: Explainability and Evaluation Anomaly Detection / Trustworthy AI
An open-source research project investigating anomaly detection baselines, explainability methods, faithfulness metrics, reproducibility, and trustworthy model evaluation.
NIDS-SCNN: Lightweight and Explainable Network Intrusion Detection Cybersecurity / Explainable AI
Lightweight CNN-based network intrusion detection using time-frequency representations and explainable AI.