biosignal-processing

Elata SDK is the cross-platform biosignal toolkit for building neurotechnology and remote biosensing apps on web and native.

Real-time emotion recognition with 40-channel EEG, facial analysis & PPG fusion - PyQt6 interface with DEAP dataset, KNN/SVM classifiers

A flow matching framework for EMG generation

This repository offers a pipeline for classifying insomnia using EEG, EMG, EOG, and ECG signals, featuring early and late fusion, signal preprocessing, feature extraction, and machine learning models for accurate detection.

This project demonstrates how electromyographic (EMG) signals can be used to control a prosthetic hand. The software acquires raw EMG data from surface electrodes, processes it in real time using digital filters, and translates the extracted muscle activity into control commands for servo motors.

First‑year BTech Electrical Engineering project (2020–21): NI Multisim simulation of a wearable stress‑meter with sensor‑fusion analytics.

Production-grade ensemble framework combining XGBoost, PyTorch & Sklearn - 70%+ test coverage with Optuna optimization for time-series prediction

Implementation of the core Adaptive Chirplet Transform (ACT) algorithm using THRML (Thermodynamic sampling) to efficiently find the best matching atom in the continuous space

EEG/EMG biosignal classification pipeline for prosthetic arm intent recognition, combining Classic ML, CNN, and RNN/BRNN models with shared preprocessing and demo tools.

Deterministic ECG codec — Python + Rust, CI parity-gated. Bounded clinical-mode contract: PRD ≤ 2.32% on MIT-BIH (48/48, mean PRD 1.12%); PTB-XL boundary disclosed (max PRD 5.29%). Cardiologist-equivalence and regulatory closure out of scope.

Real-time, privacy-first multimodal stress & drowsiness detection from a webcam: MediaPipe facial-AU + rPPG fusion, an 11-D RF+ExtraTrees ensemble (98.5% acc / 99.95% ROC-AUC, 5-fold CV), with a full ablation study and significance testing.

ECG signal processing pipeline for Atrial Fibrillation detection and Heart Rate Variability (HRV) analysis using the MIT-BIH Atrial Fibrillation Database.

ECG preprocessing, R-peak detection and QRS analysis using MATLAB.

R functions for processing pupillary light reflex (PLR) recordings — artifact removal, feature extraction, and batch analysis

Physiological stress detection using ECG, EDA, and respiration features from WESAD, evaluated with LOSO cross-validation.

Multimodal EEG-EMG deep learning pipeline for upper-limb movement decoding - 84.1% accuracy across 37 subjects (NeBULA dataset)

Operator-based framework for interpretable early-warning detection in coupled EEG/ECG signals using phase embeddings and deterministic instability gates. Supports ablations, synthetic validation, and reproducible biosignal analysis.

Real-time muscle fatigue monitoring using surface EMG, Arduino Uno, and threshold-based fatigue detection for wearable biomedical applications.

A biomedical signal classifier that combines ECG with a second signal (PPG or accelerometer) in a PyTorch model, served via FastAPI, featuring a Next.js visualization and a Claude API layer that generates a clinical-style explanatory report on detected anomalies (including explicit disclaimer stating that this is a project not a diagnostic tool).

Machine learning pipeline for detecting cognitive fatigue from EEG biosignals with 85%+ F1-score using ensemble methods