Highlights and selected projects. Each section can later be expanded into a dedicated page with figures, videos, datasets, and downloads. Detailed legacy material is linked from the Archive.
Six curated highlights. These anchors are referenced from the homepage.
Efficient clustering designed for very large spatial datasets, focusing on practical runtime, stability, and strong empirical validation.
Modular learning architecture where pre-trained neural networks operate as higher-order “neurons” inside a larger adaptive system—enabling scalable problem solving.
Probabilistic neuron behavior inspired by biological firing, combined with learning rules and implementation thinking suitable for efficient hardware.
Low-cost 3D imaging approach using focus stacking and intelligent focus/defocus detection for thick, irregular surfaces—useful in materials science and biology.
A suite of planetary visualization and analysis tools, including crater-related workflows and 3D surface exploration methods.
Improving weak-signal detection in noisy environments using noise signature discrimination and anti-coincidence concepts, with safe emulator-based testing options.
Selected algorithmic work across neural networks, prediction, recognition, and optimization.
Sparse associative memory approach enabling large-scale content-addressable memory with optimized partial connectivity.
Tool-building for rapid experimentation: neural network toolbox (C++), pattern recognition toolbox, and interactive design utilities.
Time delay neural models and predictive algorithms applied to real-world series such as rainfall.
Algorithms aimed at restoring degraded images, including research-driven optimization and evaluation.
End-to-end recognition pipeline with assistive outputs for accessibility and usability.
3D face reconstruction from limited views using reduced landmarks and practical geometric constraints.
Systems that combine sensing, electronics, and AI for real measurement environments.
Reflective light microscopy + focus stacking + intelligent sharpness detection for 3D reconstruction.
Signal emulator tools supporting safe development of radiation detection methods without live sources.
Embedded and FPGA-oriented modules intended for payload-class constraints and reliability needs.
Practical measurement and logging systems for laboratory signals and field deployment.
Visualization, analysis, and classification workflows for planetary surface data.
Rock/material classification workflows using computational intelligence methods.
Modeling tools and analysis workflows supporting crater research and classification.
Interactive 3D exploration and stereo/anaglyph browsing tools for lunar/Mars surfaces.
AI-assisted analysis of biological sequence data and database-scale mining utilities.
Knowledge base + neural network approach for prediction and interpretation.
Tools for extracting properties, mining datasets, and supporting sequence alignment workflows.
Low-power sensing, local analytics, and distributed system ideas for practical deployments.
Hybrid and solar-powered sensor network projects for smart-city and irrigation-scale applications.
Power-aware measurement and processing pipelines (e.g., RMS, filtering, compression) on microcontrollers.
Robots with grippers and navigation systems, and neural-model-based predictive robot control concepts.
