noise-filtering

Mad Location Manager is a library for GPS and Accelerometer data "fusion" with Kalman filter

Javascript based Kalman filter for 1D data

Simple Python and Julia implementations of the 1€ Filter. The codes can be used as a pseudocode for implementing the algorithm in other languages.

An individual project related to denoising for event camera

Full Stack Development of Autonomous Navigation of a self Balancing Segway Robot.

GenPark AI Agent Skill - Information entropy and lexical diversity turn pruner stripping conversational noise while preserving core user intent.

GenPark AI Agent Skill - Information entropy and lexical diversity turn pruner stripping conversational noise while preserving core user intent.

Filtering unwanted background noise from .wav files using different algorithms (Moving Average, Frequency Domain Filter and Spectral Subtraction)

The noise generator is a simple plugin which generates either white, pink or brown noise.

ClearSpeak is a real-time audio transcription application using Google's Speech-to-Text API. It features a Tkinter-based GUI, filtering background noise, and providing clear speech transcription.

Noise-Adaptive Driving Assistance System (NADAS) using Deep Reinforcement Learning, State-Estimation & State Representation

Detecting cancerous lesions by implementing a segmentation method based on histogram thresholding and color space optimization.

Spatial operations use pixels in a neighborhood to determine the present pixel value. Applications include filtering and sharpening.

Codebase of the conference paper: Assessing Adversarial Effects of Noise in Missing Data Imputation

End-to-end pipeline for hydrometeorological time-series processing and flash-flood modeling with a ML framework. Integrates rain-gauge and river-stage data through advanced preprocessing (with outlier detection & noise filtering) for flood forecasting with classical ML models; performance evaluation and exploratory/statistical analysis of results.

The `KalmanFilter` class implements the Kalman Filter algorithm for estimating the state of linear dynamic systems using noisy measurements. The class accepts system matrices, initial state, and covariance, and provides `predict` and `update` methods for state prediction and refinement based on new observations.

Smart human voice recorder using Silero VAD (PyTorch) + frequency analysis — detects & records speech while filtering background noise. Auto-stops on silence with configurable thresholds.

Indoor Human Walking Path Reconstruction from a FMWC Radar Signal

A threshold-counting node for Node-RED: fires only after a configured number of trigger messages arrive within a rolling time window — suppressing false positives from noisy or flapping sensors. Latching output with cooldown, crash-surviving persistence, and every decision, including suppressed ones, observable on a dedicated events output.

MATLAB simulations covering Fourier transforms, Nyquist sampling limits, quantization noise, and BPSK/QPSK modulation.