Related Experiment Video
Updated: Aug 29, 2025
05:48
Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
Published on: August 9, 2024
1.6K
Regional Language Speech Recognition from Bone-Conducted Speech Signals through Different Deep Learning Architectures
Venkata Subbaiah Putta1, A Selwin Mich Priyadharson1, Venkatesa Prabhu Sundramurthy2
1Department of Electronics and Communication Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Avadi, Chennai, India.
Computational Intelligence and Neuroscience
|September 5, 2022
Summary
Bone-conduction microphones (BCMs) offer noise resistance but can reduce speech quality. Evaluating different skull locations, the larynx proved optimal for BCM speech intelligibility, enhancing recognition accuracy.
Area of Science:
- Speech processing
- Biomedical engineering
- Acoustics
Background:
- Bone-conduction microphones (BCMs) capture speech via skull vibrations, offering superior noise resistance compared to air-conduction microphones (ACMs).
- BCMs exhibit a distinct frequency response, focusing on lower speech frequencies, which can impact speech quality and intelligibility.
- Current automatic speech recognition (ASR) systems struggle with BCM data due to signal characteristics and the inability to model phonemes effectively.
Purpose of the Study:
- To investigate the impact of different bone locations on the speech intelligibility of BCM signals.
- To evaluate the performance of listener and deep learning models in recognizing BCM-acquired speech.
- To determine the optimal bone location for BCM use in speech recognition applications.
Main Methods:
- BCM speech signals were acquired from Tamil words using three distinct bone locations: right ramus, larynx, and right mastoid.
- Speech intelligibility was assessed using both human listener evaluations and advanced deep learning architectures (CapsuleNet, UNet, S-Net).
- Phoneme modeling and signal characteristics were analyzed in relation to bone conduction points.
Main Results:
- The larynx bone location demonstrated significantly improved speech intelligibility compared to the right ramus and right mastoid.
- Deep learning models, including CapsuleNet, UNet, and S-Net, corroborated listener findings, validating the superiority of the larynx site.
- Analysis revealed that the specific bone location critically influences the fidelity of captured speech signals and phoneme representation.
Conclusions:
- The larynx is the optimal bone location for acquiring BCM signals to maximize speech intelligibility.
- Advanced deep learning techniques can effectively model and recognize BCM speech, with performance contingent on signal acquisition site.
- Further research into BCM signal processing and ASR integration can leverage these findings for improved noisy speech recognition systems.
More Related Videos
Related Concept Videos
01:18Classification of Bones
6.3K
The bones of the human skeletal system are of varied shapes, sizes, and functions. They can be classified based on their shape and function into four major classes: long bones, short bones, flat bones, and irregular bones. Some classifications include a fifth type, the sesamoid bones, as a separate class, whereas others categorize them under short bones.
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The...
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The...
6.3K
01:12Regional Terms
11.5K
Regional terms describe anatomy by dividing the body parts into different regions that contain structures involved in contributing similar functions. Using these terms helps increase the accurate description and identification of the particular region of interest or region affected by the disease.
Primarily, the human body has two major regions, the axial and appendicular regions. The axial region comprises regions from the head to the abdomen and makes up the central body axis. In contrast,...
Primarily, the human body has two major regions, the axial and appendicular regions. The axial region comprises regions from the head to the abdomen and makes up the central body axis. In contrast,...
11.5K