Fast R-CNN: Everything you need to know from the paper

[@Denverse]

At 3:00 I think it needs a correction. You mentioned the selective search runs on the feature maps/images, where as the selective search actually runs on input original image, and each of the proposed regions from original image maps on to the feature maps produced by backbone CNN.

[@chandank5266]

yeah exactly, I was also having the same doubt

[@amirparsa_s]

I'm just waiting for the faster R-CNN video. Keep going ;)

[@soroushmehraban]

Thanks! Appreciate your support.

[@rishidixit7939]

Is selective search applied and then the Image is fed or the Image is fed then selective search is applied on the last layer ?

[@kevinjoythomas6528]

Great video thank you

[@gamerfawaz1234]

Hi going through your CV videos , these are amusing, have used proper visualization to explain each step, which I am looking for to refresh my knowledge, keep shining.

[@soroushmehraban]

Thanks for the feedback!

[@ericsy78]

Thank you for sharing your knowledge

[@amortalbeing]

Really loved the SVD part. Could you have videos on these as well and explain them all. Those were very good.
Thanks a lot.

[@soroushmehraban]

Thanks for the feedback. Do you mean videos more related to linear algebra? It's not really my thing actually. But I'll certainly try creating a playlist if I see many people want that.

[@amortalbeing]

@@soroushmehraban yeah to some extend. The thing is SVD and the details you went through no body goes through. I actually learned for the first time what they Ctually meant for and why they are used. It was really great and very intuitive. I'd really wanted to hear more on that but you said that sufficed. I guess since you elaborated it using a real world example it just made it click and 10x more intuitive.
So I guess if you could continue on explaining the same way as you just did, things will be great.
Really loved your explanation. Great job.

[@soroushmehraban]

@@amortalbeing Thanks 😀If you're interested to learn more, I strongly recommend watching Steve Brunton's videos about SVD. I watched that myself and, quite frankly, I believe it's even more intuitive.

[@amortalbeing]

@@soroushmehraban thanks alot but I've never heard of him. What's his channels name?

[@soroushmehraban]

@@amortalbeing His name (Steve Brunton). I believe if you search SVD on TH-cam, his video comes at the top.

[@罗培祥]

Cool video！I learned a lot from it. Thank you very much for your work

[@soroushmehraban]

Glad you liked it!

[@pranavgandhiprojects]

WOw so well explained....thankyou very much:)

[@AlexXPandian]

Is truncated SVD the same as PCA ?

[@Javad-ek4es]

Nice job! May you please give me some link to access your slides, too?

[@soroushmehraban]

Thanks Javad
Could you send me an email so that I can send it through email? You can find my email address by clicking on "More about this channel" under my username on my homepage.

[@nirmalthumbar]

This is really good and easy to understand! What topic are you considering for your next video?

[@soroushmehraban]

Thanks, Nirmal! It's going to be Faster R-CNN, for sure.

[@alinaderiparizi7193]

👍 that was awesome.

[@soroushmehraban]

Glad you liked it!

[@JalenBrunsonBurner]

Great Video!

[@soroushmehraban]

Thanks, Zeke!

[@siojurnalispipin]

Nice video. Thanks for sharing.

[@soroushmehraban]

Thanks, Sio.

[@vemulasuman6995]

Excellent

[@JosephLin117]

good!

[@soroushmehraban]

Thanks!

[@alianariio]

Great video!

[@soroushmehraban]

Glad you liked it, Ali!