Al right folks. Finally wrote a piece of Artificaial Neural Network for Jarvis.
Jarvis officially has the brain cell.
Of course that is not the end of it nor is it the complete building block collection. However I am pleased to announce that JARVIS can finally generalize stuff. Also He has ears and a vocal box. Just kidding. He just has a sound class to communicate with the speaker and mic.
No sense is made of the sound yet however. Inputs and outputs are working fine. I need to figure out a model for the sound system. Just read about something called Language deaf. I think Ive experienced it a bit. I went to Thailand a few years back as a student. The hotel we stayed at had a manager who also spoke hindi. He asked me my name in hindi(my mother tongue) and i told him i did not speak Thai. I was looking for thai in the words he spoke. I think that is what is called being language deaf.
If the language deaf thing is correct( I have to read up on that. Dont even know if that is what it is called.) it means that our brain is constantly hearing sounds but is actually comparing them to models in our mind to figure out what they sound like. So in fact it is a massively parallel process as we have a lot of sounds that we can recognize on an average. I will work on that later.
Meanwhile im working on sight and text via keyboard too. Most probably I will reorganize the directory structure based on the brain(Need to read that too.)
Thats all for the update.
In case you are looking to help. Here is the git repo
Also the references for neural networks that i used to build the network for jarvis:
Jarvis officially has the brain cell.
Of course that is not the end of it nor is it the complete building block collection. However I am pleased to announce that JARVIS can finally generalize stuff. Also He has ears and a vocal box. Just kidding. He just has a sound class to communicate with the speaker and mic.
No sense is made of the sound yet however. Inputs and outputs are working fine. I need to figure out a model for the sound system. Just read about something called Language deaf. I think Ive experienced it a bit. I went to Thailand a few years back as a student. The hotel we stayed at had a manager who also spoke hindi. He asked me my name in hindi(my mother tongue) and i told him i did not speak Thai. I was looking for thai in the words he spoke. I think that is what is called being language deaf.
If the language deaf thing is correct( I have to read up on that. Dont even know if that is what it is called.) it means that our brain is constantly hearing sounds but is actually comparing them to models in our mind to figure out what they sound like. So in fact it is a massively parallel process as we have a lot of sounds that we can recognize on an average. I will work on that later.
Meanwhile im working on sight and text via keyboard too. Most probably I will reorganize the directory structure based on the brain(Need to read that too.)
Thats all for the update.
In case you are looking to help. Here is the git repo
Also the references for neural networks that i used to build the network for jarvis:
No comments:
Post a Comment