Adaptive Whitening Preprocessing Applied to Onset Detectors , Mirex 2007
نویسندگان
چکیده
Adaptive whitening is a preprocessing technique for STFTbased onset detectors, which can improve the performance of onset detectors by compensating for spectral rolloff and dynamic variability [7]. In this MIREX submission we applied adaptive whitening to a selection of STFT-based onset detection functions (ODFs) from recent literature. Our code is oriented primarily towards use in real-time systems, rather than offline analysis – hence some of the design choices noted herein.
منابع مشابه
Adaptive whitening for Improved Real-Time audio onset Detection
We describe a new method for preprocessing STFT phasevocoder frames for improved performance in real-time onset detection, which we term “adaptive whitening”. The procedure involves normalising the magnitude of each bin according to a recent maximum value for that bin, with the aim of allowing each bin to achieve a similar dynamic range over time, which helps to mitigate against the influence o...
متن کاملMultiple Fundamental Frequency Estimation Using Spectral Structure and Temporal Evolution Rules
This paper describes a method submitted for the MIREX 2010 Multiple Fundamental Frequency Estimation & Tracking Task 1, which uses pitch candidate selection rules employing spectral structure and temporal evolution. For preprocessing, the Resonator Time-Frequency Image of the input signal is employed as a time-frequency representation, a noise suppression model is used, and a spectral whitening...
متن کاملTwo Nonnegative Matrix Factorization Methods for Polyphonic Pitch Transcription
Polyphonic pitch transcription consists of estimating the onset time, duration and pitch of each note within a music signal. Adaptive signal models such as Nonnegative Matrix Factorization (NMF) appear well suited to this task, since they can provide a meaningful representation whatever instruments are playing. In this paper, we propose a simple transcription method using minimum residual loudn...
متن کاملAdaptive Least Squares Image Restoration Using Whitening Filters of Short Length
This paper presents a regularized least squares algorithm for the restoration and reconstruction of images. A whitening filter of short length provides the regularization function. An adaptive version of the algorithm is developed by matching a weighting function to the regularization function. The adaptive regularization simultaneously leads to proper noise suppression and enhanced resolution ...
متن کاملBest Practices for Convolutional Neural Networks Applied to Object Recognition in Images
This research project studies the impact of convolutional neural networks (CNN) in image classification tasks. We explore different architectures and training configurations with the use of ReLUs, Nesterov's accelerated gradient, dropout and maxout networks. We work with the CIFAR-10 [15] dataset as part of a Kaggle competition [8] to identify objects in images. Initial results show that CNNs o...
متن کامل