کنفرانس مهندسی برق ایران

صفحه اصلی / سی و یکمین کنفرانس بین المللی مهندسی برق

An Iterative Post-processing Method for Speech Source Separation in Realistic Scenarios

نویسندگان :

Iman Shahriari¹ Hossein Zeinali²

1- Amirkabir University of Technology (Tehran Polytechnic) 2- Amirkabir University of Technology (Tehran Polytechnic)

کلمات کلیدی :

Speech Source Separation،Speaker Embedding،Deep Learning

چکیده :

The purpose of this paper is to design a speaker-independent Blind Source Separation (BSS) system which aims to reduce the word error rate (WER) metric on Persian speech data. The main idea behind this method is that it tries to improve the quality of the output of any baseline separation system with the help of an iterative model by removing the remained parts of the interferer speaker from each source. For this purpose, we use embedded representations of the input speech signals. In addition, our system benefits from a convergence metric that aims to purify the output signals. To evaluate the proposed method, we have collected a dataset that contains about one hour of real phone calls from landline phones. Although most of the energy of some consonant phonemes appears in high-frequency bins which are filtered in telephony speeches, our method can handle this condition by properly removing the interference. Experimental results based on different metrics have proved the effectiveness of the proposed method.

لیست مقالات

لیست مقالات بایگانی شده

Selecting the Economical Energy Storage System for Photovoltaic Power Plants

Zahra Moradi-Shahrbabak

بررسی یک روش معکوس برای استخراج ثابت دی الکتریک محلی با استفاده از میکروسکوپ نوری روبشی میدان نزدیک

علی اقراری - محمد نشاط

Investigating Validity and Reliability of The Features Extracted by a 5R Vertical Robot for Arm Motion and Learning Assessment

Sarvenaz Bourbour - Fariba Bahrami Boodelalou - Ghorban Taghizadeh

A Consensus-Based Approach for Short-Circuit Fault Type Detection in DC Microgrids Using ANFIS

Mohammadreza Mohammadhasani - Javad Sadeh

A Novel UAV-enabled V2V Mobile Network: A Reinforcement Learning Approach

Hossein Mohammadi Firouzjaei - Javad Zeraatkar - Mehrdad Ardebilipour

A Low-Power Cyclic Vernier Time-to-Digital Converter for In Pixel Applications

Mohammad Hasan Pass - Sayed Masoud Sayedi - Seyed Amir Reza Ahmadi Mehr

Holographic Technique Inspired Multi-Beam Cylindrical Leaky-Wave Antenna

Mohammad Amin Chaychi Zadeh - Nader Komjani - Sajjad Zohrevand

Depth Estimation in Monocular Images of Inside the Sewer Pipes for Navigation of Inspection Robots

Zeinab Maroufi - Alireza Hadi Hosseinabadi - Reza Askari moghadam

Backstepping-based Adaptive Constrained Control of Passive Torque Simulator Using Function Approximation Technique

Seyyed Amirhossein Saadat - Mohammad Mehdi Fateh - Javad Keighobadi

کنترل سطح آب درام بویلر های نیروگاهی با استفاده از الگوریتم کنترل تطبیقی جدید با در نظر گرفتن کارایی و افزایش عمر شیر کنترلی (مطالعه موردی کنترل سطح آب درام IP نیروگاه سیکل ترکیبی قم)

کیوان زاهدی - محمدحسین هاشمی - محسن منتظری

بیشتر

ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 42.8.0