Trjtdtk

Why does processed meat contain preservatives, while canned fish needs not?

Is it cheaper to drop cargo drop than to land it?

Is thermodynamics only applicable to systems in equilibrium?

What is the strongest case that can be made in favour of the UK regaining some control over fishing policy after Brexit?

In the time of the mishna, were there Jewish cities without courts?

Why was Germany not as successful as other Europeans in establishing overseas colonies?

Do I have to worry about players making “bad” choices on level up?

Why is current rating for multicore cable lower than single core with the same cross section?

Modify locally tikzset

If Earth is tilted, why is Polaris always above the same spot?

Packing rectangles: Does rotation ever help?

A non-technological, repeating, visible object in the sky, holding its position in the sky for hours

What's the polite way to say "I need to urinate"?

Minimum value of 4 digit number divided by sum of its digits

Any examples of headwear for races with animal ears?

Historically, were women trained for obligatory wars? Or did they serve some other military function?

How to creep the reader out with what seems like a normal person?

Why do computer-science majors learn calculus?

Need help understanding harmonic series and intervals

What is the difference between `a[bc]d` (brackets) and `ab,cd` (braces)?

What word means to make something obsolete?

Were there two appearances of Stan Lee?

When and why did journal article titles become descriptive, rather than creatively allusive?

How to set the font color of quantity objects (Version 11.3 vs version 12)?

How is R-FCN faster than FRCNN?

How is H2O faster than R or SAS?How do subsequent convolution layers work?Is one big network faster than several small ones?How to make a CNN predict a continuous value?Faster-RCNN how anchor work with slider in RPN layer?Backpropagation in Faster R-CNNFaster R-CNN wrapper for the number of RPNs in the layer dimensions?How is Stochastic Gradient Descent done in Faster RCNN?How to train two neural networks togetherClassify 3D Gesture

0

$begingroup$

During my first pass through the architecture for R-FCN, I believed that it was because their region proposal network generated regions of interest without performing the "small network slide" over every possible anchor (the same way that FRCNN performs.)

Now that I'm looking into implementations and attempting to replicate the results myself, I'm having trouble understanding how the R-FCN is faster than the FRCNN if their region proposal networks are the same.

Am I misunderstanding these architectures? I understand the idea behind k^2 feature maps but I'm not understanding how it speeds up the computation in practice.

Thanks!

machine-learning neural-network cnn convolution faster-rcnn

share|improve this question

asked Apr 8 at 17:45

user1519665user1519665

11

$endgroup$

add a comment | 

0

$begingroup$

During my first pass through the architecture for R-FCN, I believed that it was because their region proposal network generated regions of interest without performing the "small network slide" over every possible anchor (the same way that FRCNN performs.)

Now that I'm looking into implementations and attempting to replicate the results myself, I'm having trouble understanding how the R-FCN is faster than the FRCNN if their region proposal networks are the same.

Am I misunderstanding these architectures? I understand the idea behind k^2 feature maps but I'm not understanding how it speeds up the computation in practice.

Thanks!

machine-learning neural-network cnn convolution faster-rcnn

share|improve this question

asked Apr 8 at 17:45

user1519665user1519665

11

$endgroup$

add a comment | 

$begingroup$

During my first pass through the architecture for R-FCN, I believed that it was because their region proposal network generated regions of interest without performing the "small network slide" over every possible anchor (the same way that FRCNN performs.)

Now that I'm looking into implementations and attempting to replicate the results myself, I'm having trouble understanding how the R-FCN is faster than the FRCNN if their region proposal networks are the same.

Am I misunderstanding these architectures? I understand the idea behind k^2 feature maps but I'm not understanding how it speeds up the computation in practice.

Thanks!

machine-learning neural-network cnn convolution faster-rcnn

share|improve this question

asked Apr 8 at 17:45

user1519665user1519665

11

$endgroup$

share|improve this question

asked Apr 8 at 17:45

user1519665user1519665

11

share|improve this question

asked Apr 8 at 17:45

user1519665user1519665

11

asked Apr 8 at 17:45

user1519665user1519665

11

asked Apr 8 at 17:45

user1519665user1519665

11