The article explains how the Kalman filter fuses noisy sensor measurements with model predictions using Bayesian updates, demonstrates the process with intuitive 1‑D and 2‑D Python examples, and shows its real‑world applications such as navigation, autonomous driving, and signal processing.
This article explains how Douyin leverages large language models to build an end‑to‑end user‑experience pipeline that detects signals, understands feedback, attributes issues, and automates governance, turning reactive fixes into proactive, data‑driven product improvements.
This tutorial walks through STM32F103ZET6 ADC fundamentals, explains key parameters, demonstrates how to use the VOFA+ PC tool for data visualization, and provides step‑by‑step implementations of several filtering algorithms—including first‑order complementary, median, arithmetic average, moving average, limit‑average, and Kalman filters—complete with code examples and performance analysis.
This guide provides a thorough overview of essential analog and digital circuit concepts, covering basic components, common circuit topologies, power regulation methods, signal‑conditioning techniques, and interface standards, enabling readers to build a solid foundation for designing and analyzing electronic systems.
Noise in time‑series data hampers analysis, so this article systematically examines six widely used smoothing techniques—moving average, exponential moving average, Savitzky‑Golay, LOESS, Gaussian filter, and Kalman filter—detailing their principles, key parameters, performance traits, suitable scenarios, and a quantitative RPR evaluation metric.
Fourier series, introduced by Joseph Fourier to solve the heat equation, uses sine and cosine functions to represent complex phenomena, enabling the decomposition of problems into simpler parts and finding applications across wave motion, physics, finance, and modern digital signal processing.
This article explains the mathematical definition of convolution, walks through discrete and continuous examples, demonstrates its use in signal smoothing with moving averages, and surveys its wide-ranging applications in signal processing, communications, computer vision, seismology, medical imaging, and statistics.
This article introduces Fourier series, explains how any musical note can be decomposed into its constituent frequencies, demonstrates generating a composite waveform with Python, computes its Fourier coefficients, visualizes the results, and discusses practical applications such as audio analysis, instrument classification, and digital music compression.
At ICASSP 2023, Kuaishou’s audio team presented a low‑latency hierarchical fusion network for full‑band acoustic echo cancellation, detailing its multi‑stage design, asymmetric windowing, loss functions, training strategy, and achieving second place in the non‑personalized AEC Challenge, with real‑world deployment results.
The article demonstrates how to boost iOS chat session page performance by refactoring heavy data‑source and business‑data processing into ReactiveObjC signals, using combineLatest and flattenMap to shift intensive tasks off the main thread, improve readability, and eliminate frame drops.
The article provides a comprehensive overview of digital watermarking, covering its fundamental concepts, security features, embedding/detection/extraction processes, major algorithm families, practical applications such as copyright protection and anti‑counterfeiting, and future research directions in multimedia information security.
MicNest, the Best Paper Runner‑Up at ACM SenSys 2023, introduces a long‑range instant acoustic localization system that lets drones emit PRN‑modulated sound pulses captured by a four‑mic ground array to compute ToA/TDoA, achieving sub‑meter (0.5 %) accuracy up to 120 m in noisy urban settings and enabling precise, robust landing for Meituan’s delivery fleet.
The article explains the origin, mathematical basis, and practical impact of the Discrete Cosine Transform (DCT), showing how this 1970s algorithm became the core of JPEG, MPEG and modern video codecs, and recounts the overlooked story of its inventors.
Originating from three largely unknown Indian engineers in the early 1970s, the Discrete Cosine Transform (DCT) became the cornerstone of modern image and video compression standards such as JPEG and MPEG, enabling efficient lossy compression by converting spatial data to frequency components and exploiting human visual perception.
Adaptive filtering predicts time series by iteratively adjusting weight coefficients based on prediction errors, using a learning constant to control adaptation speed, making it a simple yet powerful method that leverages all historical data for continuously improving forecasts.
This article presents a non‑reference audio quality assessment method that leverages a bidirectional LSTM network to predict perceptual scores from spectral features extracted via FFT, describing the system workflow, technical advantages, data preparation, loss design, and TensorFlow implementation details.
This article presents a no‑reference audio quality assessment system that leverages a bidirectional LSTM network to extract spectral features via FFT and predict perceptual scores, describing the architecture, technical advantages, data preparation, loss design, and TensorFlow implementation.
This article presents an automated neural‑network framework based on a deep Siamese architecture to learn similarity representations between ECG recordings, covering ECG fundamentals, exploratory data analysis, signal preprocessing, model construction with Keras, and demonstrates how the trained network yields similarity scores applicable to broader signal‑matching tasks.
This article explains NetEQ, the adaptive audio jitter buffer and packet‑loss concealment technology used in WebRTC, detailing its architecture, core modules, jitter‑estimation algorithms, decision logic, and speed‑change processing, and shows how a custom server‑side implementation can improve stability and listening experience.
This article details a comprehensive redesign of acoustic echo cancellation, automatic gain control, and automatic noise control for real‑time communication, combining WebRTC and Speex to improve delay estimation, linear filtering, and non‑linear processing, and demonstrates superior performance over the original WebRTC solution.
This article introduces ten foundational algorithms—from sorting and Fourier transforms to encryption and graph analysis—explaining their core principles, historical origins, and the pivotal roles they play in today’s software, security, and data processing systems.
This article explains how to use the Hilbert transform to obtain signal feature values, defines R and J envelope statistics and bispectrum features, and provides Python code that generates signals, adds noise, computes these features, and visualizes their variation with amplitude, frequency, and phase.
The article explains Android audio visualization by decoding PCM data, applying Fourier and Fast Fourier transforms to obtain frequency spectra, comparing the built‑in Visualizer API with a custom JNI‑based implementation, and detailing FFT processing, smoothing, buffering, and Canvas rendering techniques for smooth, performant visual effects.
This article explores the acoustic foundations of audio compression, detailing how human perception models enable non‑uniform quantization in G711 (PCMA/PCMU) and how MPEG‑1 layer‑III (MP3) leverages frequency‑domain subband analysis, masking effects, and psychoacoustic thresholds to achieve high compression ratios without sacrificing perceived audio quality.
Tencent Cloud Gaming Multimedia Engine (GME) tackles the intense, non‑stationary background noise of internet cafés by employing a specialized voice‑activity detection algorithm combined with mean removal, low‑pass and numerical filtering, enabling it to isolate a primary speaker’s voice and suppress surrounding chatter, clicks, and music for clear in‑game communication.
This article describes a C# program that continuously captures microphone audio, uses volume thresholds derived from Fourier analysis to automatically start and stop recording, and provides the full configuration and source code needed to build a smart recorder for capturing dream speech.
The article explores the history and principle of the Fourier Transform, illustrating how this mathematical prism decomposes complex waveforms into simple sine components and underpins technologies such as audio compression, image JPEG, speech recognition, scientific imaging, and even space research.
