Tag
This paper introduces an AI-enhanced method for RF interference suppression using transformer models with a Finite Scalar Quantization tokenizer to boost performance and minimize latency, validated via audio metrics in digital signal scenarios.
This report investigates reinforcement fine-tuning strategies for adapting the Qwen2.5-3B-Base model to signal mathematical problems, achieving a more than threefold improvement in accuracy over the base model.
An open-source phased-array radar project, AERIS-10, designed for DIY enthusiasts and researchers, offering 360° surveillance and detection of small aerial vehicles with detailed hardware and software specifications.
This article provides a step-by-step guide to calculating the Discrete Fourier Transform (DFT) by hand, demonstrating that it involves matrix multiplications similar to those in deep neural networks.
A beginner-friendly walkthrough of the Fast Fourier Transform using R, explaining how to interpret the fft function output, amplitude, phase, and the Nyquist limit.
This article explains the Fourier transform of power-law functions, focusing on the interesting case of 1/f noise and the symmetry between time and frequency domains.
This paper proposes NN-CLEAN, a hybrid framework that embeds a multi-head residual network into the iterative CLEAN extraction loop for efficient multipath parameter estimation in low-SNR channel modeling. It matches traditional Grid-Search CLEAN accuracy while greatly reducing computational complexity and enabling parallelization for real-time MIMO systems.
A practical guide to recursive filters (SMA, EMA, low-pass, and a tiny 1D Kalman) for smoothing noisy measurements with low latency and compute.
Kirk Borne shares a recommendation for a textbook on linear algebra for data science, machine learning, and signal processing, with links to Amazon and praise from academics.
Introduces DWT-Fusion, a training-free framework using discrete wavelet analysis of token log-probabilities for detecting LLM-generated text, achieving strong AUROC results on multiple datasets.
This paper identifies spectral drift in internal activations of neural networks during misclassifications and introduces Self-Detecting Neural Networks (SDNN) that monitors spectral dynamics to detect failures, achieving 79% AUROC on CIFAR-10, outperforming confidence-based methods by 25-30 percentage points.
Introduces (MPO)², a framework combining learned matrix product operator feature embeddings with compact polynomial weight tensors for efficient multivariate polynomial optimization, achieving improved performance over existing tensor decomposition based polynomial models.
This paper characterizes the error that arises when substituting the Hadamard transform (which diagonalizes dyadic convolution) for the DFT (which diagonalizes circular convolution), identifying universal zero-error positions, a nearly full-rank error operator, and a closed-form expression for expected error magnitude governed by a single alignment scalar.
The author reflects on the decline of specialized conferences like BMVC, ACCV, FG, ICIP, and ICASSP, noting that research communities are now concentrated in a few flagship conferences, leading to limited capacity and inconsistent reviews.
An in-depth exploration of the Fourier transform, starting from Fourier series and extending to non-periodic functions, with interactive visualizations and mathematical derivation.
This study computationally analyzes heart rate variability (HRV) indices in 40 healthy adults, finding that time-domain and nonlinear indices are normally distributed and stable, while frequency-domain indices show high variability. Recommended indices for accurate HRV representation include ApEn, IRRR, HRVi, SD2, MADRR, and rMSSD.
By recreating BBC test cards, the author provides an in-depth explanation of the engineering principles behind analog video signals and the application of the sampling theorem, demonstrating how to recover high-quality calibration signals from modern broadcasts.
Resonate is a low-latency, low-memory algorithm for perceptually relevant spectral analysis of audio signals, using resonator models with exponentially weighted moving averages.
Stanford released a complete Hidden Markov Model framework, enabling everyone to use the same technique that hedge funds like Renaissance Technologies employ to find signals through noise.
Researchers at MIT Lincoln Laboratory propose 'principle-driven foundation models' that encode signal-theoretic physical principles (Fourier decomposition, energy conservation, symmetry) instead of learning statistical correlations from large paired datasets. Trained exclusively on RF data, their 1.99M parameter frozen encoder achieves 77.7% average accuracy across 15 diverse tasks spanning audio, images, text, and video without any fine-tuning on target domains.