Sinhgad Technical Education Society’s

SINHGAD ACADEMY OF ENGINEERING

Department of Electronics and Telecommunication.

PG Students 2012-13

Seminar On Marathi Language Speech Synthesizer Using Concatenative Synthesis Strategy

Submitted By Guided by Mr. Sawane Padmakar D. PROF. Patil M.M.

Outline Introduction Speech Synthesis Process Project work Hardware Design -- Feature of Spartan-6 -- Power Supplies -- Spartan-6 Board Power Connections -- FPGA Interfacing to PS/2 keyboard. --Electrical interface --Protocol for receiving data from the keyboard --The keyboard scan-codes -- Audio (AC-97)

Software Xilinx 13.4 --Why Xilinx

-- RTL Schematic

-- Test Bench simulation Implementation Conclusion References

Introduction

Speech is one of the most vital forms of communication in

everyday life.

Human computer interaction but still rural communities lag

behind technologies.

One good solution for this problem would be computers talking

to the common man in the language he is comfortable to

communicate in.

Speech Synthesis Process

Project work divided in Two Part

Fig 2 The speech synthesis system.

1. Hardware Selection & Design2. Software Module:-• PS2 Keyboard interface with FPGA. • Concatenation & Marathi word searching in database.• Audio voice of Marathi word which is input from PS2

keyboard.

Hardware Design Features of Spartan-6 Xilinx Spartan-6 LX45 FPGA, 324-pin BGA package 128Mbyte DDR2 with 16-bit wide data 10/100/1000 Ethernet PHY on-board USB2 ports for programming and data transfer USB-UART and USB-HID port (for mouse/keyboard) two HDMI video input ports and two HDMI output ports AC-97 Codec with line-in, line-out, mic, and headphone real-time power monitors on all power rails 16Mbyte x4 SPI Flash for configuration and data storage 100MHz CMOS oscillator 48 I/O’s routed to expansion connectors GPIO includes eight LEDs, six buttons, and eight slide switches

• The Atlys board requires an external 5V, 4A.

Atlys Power Supplies

Supply Circuits Device Amps (max/typ)

3.3 VFPGA I/O, video, USB ports, clocks, ROM, audio

IC16: LT3501 3A / 900mA

2.5 VFPGA aux, VHDC, Ethernet PHY I/O, GPIO

IC15: LTC3546 1A / 400mA

1.2 V FPGA core, Ethernet PHY core IC15: LTC3546 3A / 0.8 – 1.8A

1.8 V DDR & FPGA DDR I/O IC16: LT3501 3A / 0.5 -- 1.2A

0.9 V DDR termination voltage (VTT) IC14: LTC3413 3A / 900mA

Power Supplies

Spartan-6 Board Power Connections

Fig 3 Spartan-6 Board Power Connections

FPGA Interfacing to PS/2 keyboard.

Goals: Attach a keyboard to the Spartan 6 device

through PS/2 type of port; Design a logic which can detect a character. Use VHDL during the development process.

FPGA read PS2 Data_Out & PS2 Clk_out = ‘0’So tristate buffer= high impedance mode FPGA write Logic ‘0’:- PS2 Data_Out & PS2 Clk_out = ‘1’Which pulls line lowLogic ‘1’:- PS2 Data_Out & PS2 Clk_out = ‘0’Which pulls line high

Electrical interface

Figure 4 . FPGA/Keyboard Interface

Protocol for receiving data from the keyboard

Data is received from the keyboard

Figure 5. PS/2 Protocol

Clock

Data

The keyboard scan-codes

• The keyboard sends packets of data, scan codes, to the

host indicating which key has been pressed. When a key

is pressed or held down a make code is transmitted.

When a key is released a break code is transmitted.

Every key is assigned a unique make and break code so

that the host can determine exactly what has

happened.

Scan Code SetKey Make Break Key Make Break

A 1C F0,1C N 31 F0,31

B 32 F0,32 O 44 FO,44

C 21 F0,21 P 4D F0,4D

D 23 F0,23 Q 15 F0,15

E 24 F0,24 R 2D F0,2D

F 2B F0,2B S 1B F0,1B

G 34 F0,34 T 2C F0,2C

H 33 F0,33 U 3C F0,3C

I 43 F0,43 V 2A F0,2A

J 3B F0,3B W 1D F0,1D

K 42 F0,42 X 22 F0,22

L 4B F0,4B Y 35 F0,35

M 3A F0,3A Z 1A F0,1A

Audio codec (AC-97)

Fig 6 Audio codec (AC-97)

• The Atlys board includes a National Semiconductor LM4550

AC ‘97 audio codec with four audio jacks for line-out,

headphone-out, line-in, and microphone-in .

• Audio data at up to 18 bits and 48KHz sampling is

supported, and the audio in (record) and audio out

(playback) sampling rates can be different. The microphone

jack is mono, all other jacks are stereo.

• The headphone jack is driven by the audio codec's internal

50mW amplifier.

Audio Codec LM4550

Software Xilinx 13.4 XILINX TARGETED DESIGN FORMAT ISE DESIGN SUITE SPARTAN-6 & VERTEX-6 FPGA FAMILIES XILINX IP CORE SOLUTION

RTL SCHEMATIC

Test Bench simulation of ps2interface

Implementation

The hardware layout proposed earlier is implemented on a

Xilinx Spartan-6 LX45 FPGA. Both the FSM and hardware

layout are synthesized using the Xilinx ISE 13.4 VHDL

Compiler.

Conclusion

In this paper we discussed the issues relevant to

the development of text to speech systems for

Marathi language. The Atlys circuit board is a

complete, ready-to-use digital circuit

development platform based on a Xilinx Spartan-

6 LX45 FPGA, speed grade -3. Simulation in

Xilinx 13.4 of keyboard interface with Spartan 6

using VHDL.

References Marathi Language Speech Synthesizer Using Concatenative Synthesis Strategy (Spoken in Maharashtra, India) 2010 IEEE; Atlys™ Board Reference Manual Thesis of FPGA-based implementation of concatenative speech synthesis algorithm by Praveen Kumar

Bamini VHDL: Programming by Example Douglas L. Perry Fourth Edition McGraw-Hill E. Veera Raghavendra, Srinivas Desai, B. Yegnanarayana, Alan W Black, Kishore Prahallad, “ Experiments on Unit Size for

Unit selection Speech Synthesis”, Blizzard 2008. Jerneja Zganec Gros and Mario Zganec, “An Efficient Unitselection Method for Concatenative Text-to-speech Synthesis

Systems”, Journal of Computing and Information Technology- CIT 16, 2008,1, 69-78 doi:10.2498/cit.1001049 Aniruddha Sen, “Speech Synthesis in India”, IETE Technical Review, Vol 24, No 5, Sep-Oct 2007, pp 343-350. S. P. Kawachale and J. S. Chitode, “An Optimized Soft Cutting Approach to Derive Syllables from Words in Text to Speech

Synthesizer”, in proceedings Signal and Image Processing, 2006, pp 534. S. P. Kishore and A. W. Black, “Unit size in Unit selection Speech Synthesis”, Proceedings of EUROSPEECH, Geneva,

Switzerland, 2003. P. Prathibha, A.G. Ramakrishna, R. Muralishankar, “Thirukkural II- A text-to-Speech Synthesis System” Edition, 2003 S. P. Kishore, Rohit Kumar, and Rajeev Sangal, “ A data – driven synthesis approach for Indian Languages using syllable as

basic unit,” in Proceedings of International Conference on Listeners National Language Processing (ICON), 2002. Eric Lewis and Mark Tatham, “Word and Syllable Concatenation in Text to Speech Synthesis”, Proceedings of sixth European

Conference of Speech Communication and Male E 77 23 Female F 84 16 Technology, pp615-618, ESCAs, 1999. T. Dutoit, “An Introduction to Text-to-Speech Synthesis”, Kluwer Academic Publishers, Dordrecht, 320 pp., ISBN 0-7923-

0254498-7. 1997. Paul Taylor, “Text-to-Speech Synthesis University of Cambridge

THANK YOU