Great video again. I agree the Schmitt trigger is the best way to emulate the treshhold levels in a neuron. One thing to note though. You mention that the led will keep the output low during the off-state. The circuit works, but the base of the transistor after the led is floating during that off-state. The output low-state of the schmitt trigger is 0.71V, but a red led has a threshold voltage around 1.7-2V. This in combination with the base-emitter voltage threshold would actually create a floating input at the base until around 2.4V, the way your circuit is set up at the moment. One easy fix is to put a resistor of let's say 100k from base to ground to tie the base down during the low state. You'll always want to tie "open" inputs down. Especially with complicated circuits later that produce lots of noise. Overall this is a great circuit. Keeping it simple is key. I'm looking forward to where you're going next. Great work!
@GlobalScienceNetwork
2 ай бұрын
Thank you! Yes, adding the 100K resistor would be a good addition to the circuit! I am happy to see that electrical engineers are following along and providing feedback!
@crhm99
21 сағат бұрын
I was that guy saying what is math going to do for me? Why do I have to do math? 10 years later here I am convinced that we will evolve and grow thanks to maths
@Luke-i7z
19 күн бұрын
My handsome king Mr. Wabeleski
@andrewwatts1997
2 ай бұрын
Good to see you are back ! Great stuff.
@GlobalScienceNetwork
2 ай бұрын
Thank you! I have some more great videos planned as well! Keep following along as things are going to get awesome!
@franh9833
18 күн бұрын
Most of the power of neurons really comes from how they are able to learn, of which the spiking behavior is important, they are able to learn causal behavior and correlational behavior at different spatial and temporal resolutions. It all comes down to the learning rules, you should look into STDP, inverse-STDP, dopamine-STDP, BCM rule, and dendritic computations. For ease of development simulation would be easier for testing.
@GlobalScienceNetwork
17 күн бұрын
I agree that the temporal response and timing of the neurons firing is important. In my artificial synapse video I described a few ways neurons can learn. When the network has lots of learning to do I will likely simulate the network and then hardcode the hardware based neurons initially to make it easier than hand tuning each synapse based on inputs. Starting with a simple case is key though and eventually the system needs to learn just based on its environment.
@heatedfire_
19 күн бұрын
its our famous math teacher again😁
@ScienceLab760
2 ай бұрын
Hi thanks for share and explain the design in this circuit, i want to know how integrate many units of this neurons for response with stimulus, if was possible in a future videos
@GlobalScienceNetwork
2 ай бұрын
Yeah, that is the plan for future videos! The next video is going to be how to build neural networks. What type of input and output devices we should use is an important question. Once we get a design refined on breadboards we can have the design made on PCBs which should make it easier to integrate with lots of artificial neurons and should be flexible for different input and output devices.
@haxresearch701
2 ай бұрын
good god! you are a proper master👍
@GlobalScienceNetwork
2 ай бұрын
ha ha thanks!
@RaviKodali-et6st
Ай бұрын
Sir really u r making amazing videos
@GlobalScienceNetwork
Ай бұрын
Thank you!
@keiichicom7891
14 күн бұрын
Hi, I understand the importance of understanding how to build these things with discrete components. Are more videos planned on this subject - I only saw this video and the Artificial Synapse video?
@GlobalScienceNetwork
14 күн бұрын
Yeah, the goal is to build more complex artificial neural networks all the way to non-biological life. I also have a video where I built a 4-bit computer with individual transistors you would likely enjoy.
@keiichicom7891
14 күн бұрын
@@GlobalScienceNetwork Thanks I am looking forward to it. I already watched the 4 bit computer video and enjoyed it.
@deepstories32
2 ай бұрын
i also made stuffs using arduino and coding , but your projects are just next level , it must have took a long time learning these
@GlobalScienceNetwork
2 ай бұрын
Yeah, a microcontroller is a good way to implement most robotic projects quickly. It will be interesting to see what we can build using artificial neurons!
@SHAINON117
26 күн бұрын
It's incredible please youtube ai show this to everyone please i swear most people would be interested a little bit to get a bit more awareness ❤❤❤❤❤❤❤
@GlobalScienceNetwork
26 күн бұрын
Thanks, yeah hopefully this gets shown to more people interested in AI.
@lafeo0077
11 күн бұрын
hey can we also get a video on indepth opamp applications for amplifying?
@GlobalScienceNetwork
11 күн бұрын
I was going to make a video about opamps and comparators. In most cases though I would just use a transistor as an amplifier and used the Schmitt trigger as the comparator. If my designs start to use opamps and traditional comparators I will make a video about how to build them from transistors. Opamps are useful for analog computing!
@thesimplicitylifestyle
11 күн бұрын
Genius! 😎🤖
@GlobalScienceNetwork
11 күн бұрын
Thank you!
@DavidBuzz
16 күн бұрын
Your link at the end to a hardware-based neural-network doesn't work, and links to an old 8bit video
@GlobalScienceNetwork
16 күн бұрын
Yeah, this is my most recent video, I link to my next video at the end so I will have the correct link once I get that video uploaded. Thanks for following along. I am excited to get the hardware-based neural-network video done.
@deepstories32
10 күн бұрын
its been a long time
@GlobalScienceNetwork
9 күн бұрын
Yeah I know. I took a deep dive into neural networks to make sure I make top quality videos. I am editing the next video now. Hopefully I get it up soon. Thanks for being interested!
@GameWorkflow
14 күн бұрын
Well put together videos very interesting content. You do however keep throwing out the word consciousness as if its a solved problem and throwing enough compute or neurons at it will make it go away.
@GlobalScienceNetwork
13 күн бұрын
Yeah, consciousness is something I plan to make a video about. I think there are different levels of consciousness from a simple discrete response in a logic gate being the lowest form, to then a simple analog sensor being slightly higher all the way to the consciousness of a human which is a high level of consciousness. A human is certainly not the theoretical max. How the information is flowing is also very important. For example, the Cerebellum has the majority of neurons in the human brain and there are few cases where it was missing and the individual did not even realize it. So just because there are billions of neurons firing does not inherently create consciousness. The neurons/system needs to integrate the inputs to make useful outputs.
@deepstories32
Ай бұрын
still waiting for a new video
@GlobalScienceNetwork
Ай бұрын
Thank you for being interested! I am working on the next video now. I decided to make it about the fundamentals of neural networks. I am hoping that I will get it done by Monday. I really want to get the hardware-based neural network video done but going over fundaments is important.
@deepstories32
16 күн бұрын
maybe you are taking time for research and development of your projects
@GlobalScienceNetwork
15 күн бұрын
Yeah, I did a deep dive into neural networks so I could make quality content on that subject. I am excited about my next few videos.
Пікірлер: 36