This page provides access to publicly available software tools and datasets developed at the Intelligent Embedded Systems Lab. These resources support research in embedded systems, artificial intelligence, intelligent sensing, and neural engineering, and are made available to foster transparency, reproducibility, and collaboration within the scientific community.

elasticAI.explorer

The elasticAI.explorer is a framework for hardware-aware neural architecture search (HW-NAS), automated model optimization, and hardware-adaptive deployment. It enables researchers to efficiently explore, generate, and evaluate AI models tailored to heterogeneous embedded target platforms.
Resources: GitHub

elasticAI.creator

The elasticAI.creator is an open-source framework for designing, training, and compiling neural networks optimized for FPGA-based hardware acceleration. It enables the automatic translation of trained AI models into hardware descriptions (e.g., VHDL), simplifying the development of efficient AI accelerators for embedded systems.
Resources: GitHub

elasticAI.hardware

The elasticAI.hardware provides open hardware designs and supporting components for developing intelligent embedded devices within the elasticAI ecosystem. It serves as the hardware foundation for deploying and evaluating AI applications on resource-constrained platforms.
Resources: GitHub

elasticAI.runtime

The elasticAI.runtime provides the runtime infrastructure for distributed AI applications across embedded, edge, and cloud environments. It manages communication, resource allocation, deployment, and monitoring of intelligent devices, enabling adaptive and scalable AI systems.
Resources: GitHub

denspp.offline

denspp.offline is an open-source Python framework for end-to-end processing of transient measurement data. It supports the complete workflow from raw data handling and signal preprocessing to training and evaluating deep neural networks, with a focus on reproducible embedded AI research.
Resources: GitHub

For detailed documentation, publications, and component-specific information, please visit the elasticAI.ecosystem project page. Source code and software releases are available through our GitHub organization.

Surface Vibration Gesture Recognition Dataset

This dataset was created for the development and evaluation of vibration-based gesture recognition systems on everyday surfaces. It contains recordings from four piezoelectric sensors mounted underneath a standard office desk and captures six gesture classes, including swipe, tap, and knock gestures. The dataset comprises recordings from 15 participants with a total of 9,000 annotated gesture events and supports research on signal processing, embedded machine learning, and human-computer interaction.
Resources: Dataset | Publication

Impedance Characterization Dataset for Microelectrode Arrays

This dataset supports research on impedance estimation and adaptive stimulation in neural interfaces based on microelectrode arrays (MEAs). It includes electrochemical impedance spectroscopy measurements, transient stimulation recordings, numerical finite-element simulations, and the accompanying processing code. The dataset was developed to investigate whether impedance characteristics can be reliably extracted during electrical stimulation and enables reproducible studies in neural engineering and bioelectronic systems.
Resources: Dataset | Publication