Domains

The group develops reference ‘C’ code for Audio and Speech Codecs. The reference code is well optimized for Performance and Memory to suit real time multimedia consumer applications. The reference code is available in both floating point and fixed point arithmetic which can be easily ported to any platform. The advantage of developing reference code is it greatly reduces the product development time for multiple platforms in turn reduce the cost.

The group has developed MP3, MPEG-4 AAC, WMA audio codecs and G723.1, G726 and NB-AMR speech codecs. During course of the reference codec development team has made many innovations for algorithm optimization in terms of performance and quality. The team is also involved in the innovating and developing new algorithms to improve the performance and quality of the existing algorithms. Currently the team is porting the reference code to various platforms.

      Automatic Speech recognition
It is the identification of spoken words by machine. A major challenge to the deployment of an automatic speech recognition system to consumer/automotive applications is to maintain satisfactory recognition accuracy and real-time performance for all operating conditions especially under noisy environments.

The goal of the project is to develop a flexible and scaleable Speech Recognition Engine for automotive and consumer applications. The Engine supports advanced features such as language modeling, speaker adaptation and multiple languages. The various applications include mobile phones, PDA’s, vehicle navigation systems, vehicle command and control and in-vehicle entertainment systems. The Speech Recognition Engine is being developed for various processors and Real Time Operating Systems”

      DTV Middleware:
Digital Television (DTV) Middleware is a software solution to provide digital entertainment services for broadcast TV. It acts as an intermediary among various digital television applications. Middleware provides DTV applications platform & OS independence.

DTV Middleware contains implementation of various software decoders viz tables decoders (MPEG-2, DVB, ATSC), Subtitle, Teletext, Closed Caption etc.
Advanced DTV middleware can provide digital entertainment services including video on demand, pay per view, digital music; as well as interactive IP-based applications such as web browsing and online gaming.

Toshiba TEC Corporation has three main divisions. These are Retail Information Systems Company, Document Processing & Telecommunication Systems Company, and Home Electrical Appliances group. The division at TESI works for the Document Processing & Telecommunication Systems Company which is a leading manufacturer of network-ready MFPs (Multi-Functional Peripherals).

MFPs play an important role in facilitating the flow of information and refining office operations in general. This involves a sector of technology referred to as "image information communication" and it is one that focuses on user-friendliness – allowing each and every user in a network to increase their effectiveness through easier operation, streamlined management and more efficient creation of documents. TTEC strives to add more value to their customers by providing innovative features like Out-of-the-Box Network-Ready, Print-Ready, Scan-Ready, Internet-Ready, and e-FILING (Electronic Filing).

TTEC is building a line of B&W and Colour MFPs, based on e-BRIDGE. e-BRIDGE is an all-in-one architecture that integrates print, copy, scan, and Internet Fax on a single board. The team at TESI will be working on the next generation of this architecture.

Speech and Audio Group on Mobile Multimedia
Objective of Speech & Audio group is to develop speech and audio codec’s on Toshiba’s S1, S1G and T4G platform. These codec’s are being used in Recorder and player application running on mobile multimedia chip.

Activities performed in codec development are porting of speech & audio codec on media embedded processor and audio telecom digital signal processor, implementation of platform and application specific optimization, effective utilization of memory and development of simplified external interface used by application

S1G Player:
This project aims for porting the player firmware which was developed for S1 on the S1G board.
The specification of S1G player firmware is almost similar to S1 Player firmware. This will be integrated with the S1G drivers.S1G player firmware will support the following combinations. It is a common firmware with pluggable codecs.

1. MPEG4 Video + MPEG4 AAC player
2. MPEG4 Video + AMR player
3. MPEG4 AAC stand alone player
4. MP3 stand alone player
5. H.264 Video + AMR player

The combinations 1,2,3 and 4 are already part of the S1 Player, they just need to be ported on the S1G board. But in case of 5, the H.264 video decoder still needs to be ported on S1G and integrated with the firmware with AMR decoder.

S1G Drivers:
This project aims at testing and bug fixing of S1G LCD IF, Camera IF, Scaler IF, Audio IF and Host IF drivers. Then they will be integrated with the S1G player. Now the testing/bug fixing is almost at the end.

S1 Player Application for Mobile Device
This project involves development of Player Firmware on TSB S1 hardware platform for Mobile Multimedia application. The S1 hardware comprises of MeP (Microcontroller)
+@DSP (Digital Signal Processor) driven by 125 MHz clock; along with necessary
multimedia support I/O hardware interfaces for Audio Play out, Video Play out and
Image processing (Zoom-in, Zoom-out, Rotate, Mirror and Format conversion).

A Player Firmware application will consist of following components:

a	Audio Decoder IP module Video Decoder IP module that will decode the video information present in the stream to be played out.
b	Driver IP modules that will provide necessary interface to access the respective I/O Hardware modules.
c	Control Firmware that will
	1 Schedule the above Decoder IP modules. 2 Performs necessary post processing of decoded data. 3 Renders the Audio and video output (with synchronization) that will be played on respective Output devices using the respective Driver IP modules.

A generic Open Architecture Framework (OAF) was designed in order to integrate various IP modules effectively with the Firmware module. Any new IP module (Audio/Video Decoder) designed with the above OAF guideline can be easily integrated into the Firmware module.

H264 Decoder on S1G.
The H.264/AVC video coding standard is making enormous strides in industry as the world's most advanced and sought-after coding technology, from cell phones to broadcast TV to high-definition DVD and beyond. Companies worldwide are developing technology for this standard, in both hardware and software. S1G is Toshiba’s platform and we are porting the H264Decoder on to this Platform.

The H.264 decoder on S1G will decode the H.264 encoded bitstreams conforming to H.264 Baseline Profile @level 1.2 (or lower). The H.264 decoder will be part of a larger application like a video enabled mobile phone or a digital broadcasting system. The application sits on host and the H.264 Decoder along with either G.726/ AMR or MP3 decoder shall run on S1G. H.264 decoder shall decode 15 QVGA frames/sec @256 kbps bit rate. H264 Decoder outputs the decoded data in YUV 4:2:0 format.

Scope of the project involves Memory optimization, MIPS optimization (C-level, Processor level , etc) to achieve the desired performance on S1G.

“MPEG-4 Video Encoder
Its development involves implementation of complex video encoding algorithms for perceptual modeling and encoding of visual (video) data with high degree of compression, with minimal unavoidable impact on visual quality, to enable streaming it over limited bandwidth network or storing it on limited capacity storage device, as per MPEG standards.

“MPEG-4 Video Encoders for streaming applications are supported with implementation of error-resilience techniques to enable early recovery from transmission errors. In order to maintain highest possible video quality, while avoiding overflowing or under-flowing VBV buffers at decoder end at given throughput bitrate, rate control algorithm is optimized and fine tuned to do intelligent distribution of available bit-budget across sequence based upon input conditions and data.”

H.264/MPEG-4 AVC
It is the latest and most advanced standard jointly developed by ITU and ISO/IEC. It provides a far more efficient algorithm for compressing video than any other compression method available. It outperforms all existing standards by a factor of three to four especially in comparison to MPEG-2 and provides a “network-friendly” video representation. This is emerging as the next standard for broad range of applications such as High Definition DVD, Digital TV, video conferencing, video-on-demand, multimedia messaging, Broadcast and Streaming applications.

H.264 delivers excellent quality across a wide operating range from 3G to HD and everything in between. H.264 provides exceptional performance at impressively low data rates and high quality video for mobile phone, iChat, Internet, broadcast and satellite delivery.

Digital TV Audio

Digital Consumer Audio :
Digital Consumer Audio group specializes in providing end to end audio solutions for consumer electronic products like Digital TV, DVD audio player/recorders, car infotainment products.
This group is actively involved in:

	Design and development of various audio algorithms like graphic equalizer, sample rate converter etc
	Customization of audio codecs like MP3, AAC, MLP, DTS, AC-3 etc for MeP(Media embedded Processor) based SoC(System on chip) solutions
	Audio firmware development for DVD and Digital TV Product certification for Third party technologies like DTS Decoders, Dolby Pro-logic audio post-processors, SRS labs WOW-XT/HD etc
	Codecs/Audio firmware development for emerging technologies like HD-DVD, DTV-DVD based products