Sunday, 22 July 2012

Analysis on the discussion of Voice Recognition techniques & its Vedic Principles

This was my analysis in a thread of discussion held in Foundation for Indian Scientific Heritage

रुदं द्रावयतीति रुद्र  in Vedas. The tatwa that comes to help where there is crying or sorrow is Rudra. There are lots of debates on the theory of Kalpa in Vedas. In Aghamarshana it is धाता यथा पूर्वं अकल्पयत् But not necessarily the everything repeat as it is! It will try to make everything as previous. But again previous is not explicitly said to be Kalpa. The contradiction comes when it is said that everything happens newly as in नवो नवो भवति जायमानः

For the supreme power some say it as only one, then variations as 2, 3, 4, .... 16, 24, 25, ....., 32, 33,...., 40, ...64,...72,...,84,..., 108... Sahasra, Ayuta, Laksha, Koti, .... Ananta. Its nirvivaada that with respect to the concept being explained in the quoted shloka of Brihadaaranyaka Upanishad, it is 33. But its not a Constant value. A Vedantin may see it as a seed factor which gets multiplied and may become crores. But in KarmaTa accurately wants the particular diety for the rituals. Some relate the 33 are the Devas that function in all the seven planes as said in सप्तार्ध गर्भा भुवनस्य in Atharva 9-10-17 and hence the number is 33,00,00,000. The Vedas call them 33 crores of Devas. So it is not a unique number. Atharva 10-2 asks कति देवाः कतमेत आसन् ? Then Sankyas start calculating it and gives the numbers from their Siddhanta. 

In Hiranyagarbha Sukta, every Ruk has the Vaakya “ कस्मै देवाय हविषा विधेम ” | It gives different meanings according to the context of each Ruk. So every time it gives the processes related to creation, Sustenance, and manifestation. Thereby asks who is behind all these and whom we have to worship with oblations? So it is elaborated from Yagna Sutras from Yajurveda and proper division of Devata, HavirBhaaga & its Vidhi-Vidhaana. One thing should be clearly understood that according to Vedas the Deva is nothing but the one which has the power to give. Veda recognizes them according to the properties assigned to them. It is not interpreted like jumbled reverse of Veda is Deva!

When it comes to divisions of Karma. Bhauta Shastriya analysis of Yajurveda Rudra Mantra gives 712,80,000,000 Karmas and due to these many huge karmas we also see huge Shivaghanas. So all the Akshohinis which became Bhasmibhoota in Mahabharata war were equivalent to the count of Shivaghanas. Again this multiplied with count continues to give 64152 Koti Karmas. So these many Jeeva-Prabhedas are there in this Prapancha. Rudra keeps the count of all these and Vishwakarma is the director who directs all these Karmas. These are all the positions which are occupied by the souls when they achieve the power to give. So, the Deva which has the power to give is also enormous. It is not just 33!

Let us come back to the word पूर्व in Aghamarshana Mantra. Atharva 8-10-4 there is a sutra " तां द्विमूर्धा अधोक तां माया मेवाधोक " Only human possess 2 types of Prajnas called Praapanchika Jeevana Prajna & Aatmika Prajna. अधोक तां - one has to understand the माया that has been explained in the other  mantras, Maaya Vaada, Adwaita Siddhanta, etc., and then he/she can cross the effects of Maayaa. This Maayaa makes every thing to be forgotten and sustain the living in present to continue towards future. Nobody knows about Poorva. Vedas also describe the the knowledge of this Kalpa. Kalpaantara jnana is Aspasta. So tantra also keeps a separate place for the “Aspasta Kalpaantara Jnana” as commonly witnessed in Sarvato Bhadra Yantras. So, Kaplaantara Jnana is considered as Parishita in Vedas.

In Puranas While explaining evolution, it is said that there was earwax in Vishnu’s ear. It was removed out by Brahma. Means, he did the Samyoga of Pancha-Tatwas. There emerged Madhu-Kaitabha. First shiva unified with Vishnu. Means, when the Raajasa was unified with Taamasa there emerged the earwax which is taamasa named raajasa came out. It was Rudra. Hence Panchatwa unified to emerge these Raakshasas. It goes in more depth and divides the 11 causes of creation of Jeeva Prabheda with respect to this concept as Shiva, Shambhu, Pinaki, Girisha, Hara, Sthanu, Bharga, Bhairava, Sadashiva, Kapali, Ishwara. If we just consider the Shiva part of this and see where it will be unified? Atharva 4-25-3 gives the answer:

तवव्रते निविशन्ते जनासस्त्वय्युदिते प्रेरते चित्रभानो |
युवं वायो सविता दुष्कृतमपरक्षांसि शिमिदां सेधतं

Natural & open meaning of Shiva is Shubha. Bautha Shaastra does the analysis of Shubha here. We usually think which gives profit is good. It says that what we think good in this worldly matter may be extreme bad in the future. So the first part of the mantra says about what is Laabha? Is is Shubha (vrata) for your life style? So, it has to be decided based on the type of path being chosen in leading the life. It should also be Preraka, Sahaayaka for you. Chitrabhano – Not only for you, it should be Shubha for your environment and for whole world. In the 2nd paada it is said that it should not be rejected by the basis of Jeeva - the Praana and Savita – the Jnana. Then only it will be capable to make you cross the Vrutti Maarga. This is Shubha. Not just the outlier Laabha. So the one which passes these many levels of tests will be considered as Shubha. It is called as Shiva. Then says about the combination of Shiva + Panchatwa = Which Srusti? It is elaborated in Brahmaudanam Sukta of Atharva. Then Shanbhu, Pinaki, etc., all are explained clearly.

Among the first division of species from the quadruple Shiva, Shambhu, Pinaki, Girisha, there is no concept of Rudra. रुदं द्रावयतीति रुद्र. Then there is Rudam. Where it is? There is no sorrow in first division at all. So Rudra comes into picture in the next division named Hara, Sthanu, Bharga, Bhairava. There is no death to the Jeeva, its only Rupaantara. There is an end to the positions held by Brahma, Indra, and only to the body that is born in this Earth, but not to the Jeeva.
When we want to express the Jnana, first it has to be divided based on the rule:

कल्पंतां ते दिशस्तुभ्यमापः शिवतमास्तुभ्यं भवंतु सिन्धवः |
अन्तरिक्षं शिवं तुभ्यं कल्पंतां ते दिशः सर्वाः

Creatively the letters has to be chosen based on the picture that is got by the powerful imagination. The directions of writing is purely based on the Rasas which are 9 according to linguistics. But this kind of expression is never ending. Expressing the whole knowledge in terms of words is very difficult, but if we take the Rasa, the concept and represent it with symbols, then we can represent all the possible subject matter of knowledge easily.

While expressing the knowledge in terms of audible sound, Veda says that there are 127 types of different basic Human Vaikharis. This is due to the variation of Swarocchara from the vocal cords. This is a unique trait for a person. Blind people also identify the persons based on these 127 variations of Swaroccharas. It is called as Hamsa-Swara. This is very basic property of Vaikhari of any creature. So one among them is humans’ i.e., 63+63+1. If we identify these basic vocals and make them as basic constants, then the NLP techniques can be applied to compare & categorize the incoming speech. So my suggestion while programming the speech related to Indian languages, Sanskrit in particular are as follows:

1.       Collect several samples of Vaikhari from different parts of India.
1.1.    Swarocchaara
1.2.    Vargiya Vyanjanocchaara
1.3.    Avargiya Vyanjanocchaara
1.4.    Kaagunita Ucchara
1.5.    Samkya Ucchara
1.6.    Important Dhatu Ucchara
1.7.    Vikaara Ucharaas that include fast talking, merging some of the letters while speaking, etc., which has become common among people should be considered. Here user friendliness plays very important role as the pronounciation which they are doing now cannot be changed easily.
2.       Perform DSP on them to retain clear on the “Swarocchara Vidhaana”.
3.       Identify & categorize the basic 127 variations of Swaroccharas. If this is not possible, then one has to go back to normal technique of coding the frequency of all possible words, but time consuming.
4.       Manage them in some good Object-Oriented DBMS that supports real time signal comparison, compression & retrieval in a faster manner.
5.       NLP algorithms should be chosen properly. If required, advent new algorithms to accommodate them to Indian scenario.
6.       Build the audio grammar by xml kind of technology which can interact with OO-DBMS as well as font-end in a faster manner.
7.       The end user may be given with some basic check boxes like Gender, Region of Birth, Region of residence, languages he/she can speak, etc., One of the important factors that form the Vaikhari is Water they drink, so according to Vedas we can classify mainly based on the properties assigned to 7 sacred rivers. All the other sources of water, thus follow these 7! Combinations. This is very high end research area with respect to Vedas & Technology. These will form the basis of Artificial Intelligence based Voice Recognition techniques.

This is not “Iti”. It’s just an analysis. Lots of brainstorming sessions are required for smooth continuation of this kind of research projects. There should be sufficient budget, human resource, and interests from research community, government, & public to at least use the software and appreciate the hard work behind it.

No comments:

Post a Comment