Harmonic Temporal Structured Clustering: A New Approach to CASA
نویسندگان
چکیده
あらまし 人間の聴覚機能を計算機で実現しようという試みが活発に進められており、その枠組を総称して計算論的 聴覚情景分析 (Computational Auditory Scene Analysis: CASA)と呼ぶ。近年この研究分野における興味の対象は、 Bregmanが指摘した分凝要件 [1]に基づく混合音分離法の実現にある。CASAにおける多くの従来手法の間で共通す るのは、各時刻で独立に調波成分を見つけ出すための処理 (周波数方向の群化)と、抽出された調波成分特徴量の時系 列を時間方向にスムージングする処理 (時間方向の群化)を多段処理的に行っている点である。しかしながら、より良 い群化プロセスの実践のためには、これらは本来協調し合いながら行われるべきであり、個々の音源の時間周波数全 域に渡ったスペクトル構造を一挙に推定できる方法論が不可欠であると我々は考える。本稿では、この観点から導か れる調波時間構造化クラスタリングと呼ぶ CASAのための新しいアプローチを提案する。 キーワード 計算論的聴覚情景分析, Bregmanの分凝要件, 定 Qフィルタバンク, 調波時間構造化クラスタリング
منابع مشابه
Harmonic-Temporal Factor Decomposition Incorporating Music Prior Information for Informed Monaural Source Separation
For monaural source separation two main approaches have thus far been adopted. One approach involves applying non-negative matrix factorization (NMF) to an observed magnitude spectrogram, interpreted as a non-negative matrix. The other approach is based on the concept of computational auditory scene analysis (CASA). A CASAbased approach called the “harmonic-temporal clustering (HTC)” aims to cl...
متن کاملMonaural Speech Separation Based on a 2D Processing and Harmonic Analysis
This paper proposes a new Computational Auditory Scene Analysis (CASA) approach based on a 2D spectro-temporal analysis and harmonic separation. The 2D processing, socalled Grating Compression Transform (GCT), analyzes the spectro-temporal content of the spectrogram, mimicking the processing of the primary auditory cortex. The estimated pitches from the GCT analysis are used for separation usin...
متن کاملHarmonic Temporal Structured Clustering for Multiple Fundamental Frequency Estimation
This abstract describes a method for the Multiple Fundamental Frequency (F0) Estimation and Tracking task in the Music Information Retrieval Evaluation eXchange (MIREX) 2007. The method is called Harmonic Temporal Structured Clustering (HTC), which is a kind of constrained Gaussian Mixture Model (GMM) estimation using EM algorithm. It can jointly extract F0, intensity, onset, duration of each u...
متن کاملHarmonic Temporal Structured Clustering with Unsupervised Model Learning for Multipitch Estimation
This paper describes a system for the Multiple Fundamental Frequency Estimation and Tracking task in MIREX (Music Information Retrieval Evaluation eXchange) 2009. The method is a modification of Harmonic Temporal Structured Clustering(HTC), which is a kind of constrained Gaussian Mixture Model estimation using EM algorithm. In the modification one term of a function of model power is added to o...
متن کاملHarmonic-temporal-structured Clustering via Deterministic Annealing Em Algorithm for Audio Feature Extraction
This paper proposes “harmonic-temporal structured clustering (HTC) method”, that allows simultaneous estimation of pitch, intensity, onset, duration, etc., of each underlying source in multi-stream audio signal, which we expect to be an effective feature extraction for MIR systems. STC decomposes the energy patterns diffused in timefrequency space, i.e., a time series of power spectrum, into di...
متن کامل