[attachment=15531]
BASED ON IEEE PAPER
Zero-Crossing-Based Speech Segregation and Recognition for Humanoid Robots
Sung Jun An, Rhee Man Kil, Senior Member, IEEE, and Young-Ik Ki
Project centre CRISP SYS Coimbatore.
INTRODUCTION
Robots are becoming familiar in all fields.
It allows interaction with humans and the surrounding environment.
Must have similar capacity to the human’s auditory information processing system.
Speech controlled robots provide easy access for controlling, without using wires
EXISTING SYSTEM
current automatic speech recognition (ASR) systems are not quite robust to noise
hard to attend to the selected speech source by current systems.
Difficult when multiple sound sources are present
TRANSMITING SECTION
SPEECH RECOGNITION SYSTEM
Speech Recognition System
SPEECH RECOGNITION SYSTEM
The speech recognition system is a completely assembled and easy to use
Programmable, in the sense that you train the words (or vocal utterances) you want the circuit to recognize
Features
Self-contained stand alone speech recognition circuit
User programmable
Up to 20 word vocabulary of duration two second each
Multi-lingual
Non-volatile memory back up with 3V battery onboard.
Will keep the speech recognition data in memory even after power off.
Easily interfaced to control external circuits & appliances
ENCODER WITH RF TRANSMITTER
ENCODER
HT 640 is used as encoder
Capable of encoding 18 bits of information which consists of N address bit and 18-N data bits.
Input signal to be encoded is given to AD7-AD0 input pins of encoder.
It may be from key board, parallel port, microcontroller or any interfacing device .
R F TRANSMITTER
BF 494 along with the tank circuit forms the oscillator.
When ever the high output pulse is given to base of the transistor BF 494, the transistor is conducting so tank circuit is oscillated.
Modulated signal is given LC filter section.
After the filtration the RF modulated signal is transmitted through antenna.
RECEIVING SECTION
DECODER WITH RF RECIEVER
RF RECIEVER
BC547 acts as amplifier.
Amplified signal is given to carrier demodulator section in which transistor Q1 is turn on and turn off conducting depends on the signal.
saw tooth signal which appears across capcitor is given to comparator LM558.
comparator converts the saw tooth signal to exact square pulse.
DECODER
HT648 is used as decoder.
For proper operation a pair of encoder/decoder pair with the same number of address and data format should be selected.
Decoder separate the address (A0-A9) and data signal (D0-D7).
Output data signal is given to microcontroller or any other interfacing device.
Designed to control the motor in the forward and reverse direction.
Consists of two relays named as relay1, relay2.
Relay ON and OFF is controlled by the pair of switching transistors.
DC MOTOR FORWARD REVERSE CONTROL
When relay 1 is in the ON state and relay 2 is in the OFF state, the motor is running in the forward direction.
When relay 2 is in the ON state and relay 1 is in the OFF state, the motor is running in the reverse direction.
ADVANTAGES
Input requirements are manageable
Technology and system feasibility:
Economic feasibility: Economically cheaper to implement.
Legal feasibility: Can be used in existing legal environment.
Operational feasibility: Ensures 100% of success in current implementing phase under all situations.
Less computational coimplexities
APPLICATION
Military- High-performance speech controlling robots for defense purposes.
Industry- The speech controlling robot is used for several controlling operations in Industries .
Mining- These robots can used for finding the mines .
DISADVANTAGES
It cannot be implanted in highly congested area .
RF interference will disturb the working .
CONCLUSION
We have suggested a method of speech segregation and recognition using the masking method.
For the masking method, we used the power ratio of the target to all sound sources.
Advantages of the suggested ZC-based masking methods are
The robustness to noise,
Significantly reduced
computational complexity and
3) No need to train the masks according to various spatial configurations of sound sources.