Thanks for your compliment. I’m really into it! ^^
@Indra50111
3 жыл бұрын
Thank you so much for the explanation Sir.
@quarter2018
3 жыл бұрын
You are so welcome ^^
@mohammadwahba3077
Жыл бұрын
Gooooooooooooooooooood very gooood - Go on
@quarter2018
Жыл бұрын
Thanks a lot ^^
@korilifs
3 жыл бұрын
The only video that really helped understand what is going on! Amazing! Please upload mask rcnn explanation 👍🏽👌🏻
@quarter2018
3 жыл бұрын
Thanks for your recognition. Mask R-CNN is the one in our schedule ^^.
@CaratacusAD
Жыл бұрын
Excellent video and explanation, I think this is the clearest explanation on youtube. I have a quick question. Is it possible to swap out the ROI pooling for ROI align in a Faster RCNN network. I believe it is more accurate and i've seen this proposed in a few academic papers. I'm trying to build a custom object detector based on Faster-RCNN but with additional branches to detect additional image properties like occulusion, along with the normal BBOX and classification branch. Many thanks
@quarter2018
Жыл бұрын
Since RoI pooling and RoI align are both for normalizing feature maps for further predicting, it is feasible to substitute RoI pooling with RoI align within the Faster R-CNN network. However, this substitution comes at the expense of increased computational complexity.
@deeptijoshi6118
3 жыл бұрын
Hi Huang, I really like your videos, they are so well presented and you explain them well too. Would it be possible for you to explain Mask RCNN too?
@quarter2018
3 жыл бұрын
First of all, thanks for your recognition. Now, we work on the video about Horn Schunck algorithm for optical flow estimation. Mask RCNN is the one after that. Following is the brief description about Mask R-CNN. The architecture of Mask R-CNN is almost the same as the Faster R-CNN but different in two points. (1) Mask R-CNN uses RoI align instead of RoI pooling for extracting the features of RoI. The use of RoI align makes the RoI feature extraction more accurate. About the detail of RoI align, you can refer to our video on KZitem. (2) Mask R-CNN adds a branch for mask prediction.
@brainpushups274
3 жыл бұрын
Thanks for the great lecture. How to get the ground truth d_cap and p_cap for each proposal? Do we do "proposal labeling" as the "anchor labeling" in RPN? Where are the background proposals from? Does RPN also output background proposals to RoI pooling? Thanks.
@quarter2018
3 жыл бұрын
Dear Brain Pushups: First of all, thanks for your recognition. (1) The answer is yes, we label the proposals from RPN as we do in RPN. The detail for proposal labeling can be found at the video about at 21' 36''. (2) The background proposals are those that meet this condition(𝟎.𝟏 ≤ max. IoU
@prajvalp2868
4 жыл бұрын
Thank you for the explanation Is there any code walk through available to implement the networks?
@quarter2018
3 жыл бұрын
Sorry, we still work on this and will release our code as soon as possible
@aditasukmawardani8817
3 жыл бұрын
Hii thankyou for the explanation, is there a code for this Faster R-CNN available and can be shared wih us?
@quarter2018
3 жыл бұрын
Thanks a lot! About the code for Faster R-CNN, we just use the code in the github
@aditasukmawardani8817
3 жыл бұрын
@@quarter2018 can you share the github link with us?
Пікірлер: 21