Комментарии:
great video! just to confirm...random numbers in the filter matrix is not recommended right? it's more like the example you gave right after -1,1,0 etc.
ОтветитьThank you!
ОтветитьThe best video ever. The best comment ever
ОтветитьAmazing Explanation
Ответитьgreat visual video! Well done ;)
Ответитьthanks I enjoyed it.
I was looking for this question: how should I count number of parameters in a CNN network.
this is the one of the best video to explain CNN n the filter examples are great. thanks
ОтветитьExcellent and clearly explained. Thank you for taking the time to make this video! I've been exceedingly frustrated by other content that doesn't make CNNs as intuitive as you did. Thanks again!
ОтветитьThank you from Ethiopia.
ОтветитьThank you for the clear explanation!
ОтветитьIs it similar to recursion?
ОтветитьThat was a great explanation of CNNs!
ОтветитьThis is absolutely helpful. Thank you!
ОтветитьAmazing! Thank you!
ОтветитьHey,
I have a question! Is dot product applicable to matrices? Even if it did, dot product of two 3x3 matrices is another 3x3 matrix however you have showed it as a scaler value in the conv matrix! Pleas correct me if I'm wrong in understanding this! Earliest response is highly appreciated! thanks
You guys are the bestttt...
I am following the complete playlist and so far all the concepts are so much clear.....
Continue to make video on ML
Love <3 .
great
informative
May the lord bless thank you
ОтветитьThank you
Ответитьcan someone tell me the big O notation of CNNs pleaseeee thankyouu!
ОтветитьI really like the way you teach us, it's amazing, everything was explained very precisely, thanks a lot mam!
ОтветитьOnce you can talk about machines having their own conscience, your sham maybe true.
Ответитьreally great succinct explanation
ОтветитьAmazing video, well explained
ОтветитьBORING! Julia Child was way better! Primordial Soup! 9/11!!!!
ОтветитьCNN is only useful for political propaganda 😂😂😂
ОтветитьCan you send me your email address..i will have a problm in my project
Ответитьthnks
Ответить5 years later and this video is still awesome ♥
ОтветитьFirst time i watch one of your videos videos. Insta Subscribtion. Absolutly neiled it. perfect expanation, quick and easy with examples. Gonna watch all of your videos right away
ОтветитьThank you!
ОтветитьCool
Ответитьamazing
ОтветитьIn real world image capturing for example, how do we know how many output neurons we will need? In the example with 0-9 we know it will be one of these numbers. But how do networks work there it is just a "normal" picture.
ОтветитьVery helpful. Thankyou
ОтветитьNever trust a company that wants you to pay for course content when the identities and bios of the instructors aren't made public. It's nuts, and shady as hell.
ОтветитьOK. But now how do you create the filters from datasets?
ОтветитьI got some interesting point on AI.
While Man Is able to understand other algorithms AND even design learning algorithms, I think that a network after training can make complex computations just as the human brain does AND learn really but really complex patterns.
In this Sense we are creating something Is not on our control.
Nicely explained!
ОтветитьLOVE THIS
Ответитьthe matrix dot product of 3x3 times 3x3 would also give a 3x3 matrix though, so that is I'd assume why both matrices are the same dimensions..
Oh Actually I was thinking wrong. If you have WxH filter you only get one row, and then each value less, is a value you add to the output. So 3x3 will left 2 rows and 2 columns out.
very well explained
ОтветитьFilters are what detect the patterns in the image
Their are different types of filters for detecting different types of shapes/objects
Thank you. It helped me to understand CNN )
Ответитьthis video superficially explains convolution (incomplete: why does edge detection work? what about blurring, sharpening, or different filter sizes, how do you set the weights, such that the luminosity of the image does not change) .... and how does this related to an ANN?
Ответитьwhat about rgb channels??
ОтветитьIf the first filters are initialized using random numbers, how does the more abstract filters learn to filter much more abstract concepts than differences in pixels?
Ответить