Depthwise Separable Convolution - A FASTER CONVOLUTION!

In this video, I talk about depthwise Separable Convolution - A faster method of convolution with less computation power & parameters. We mathematically prove how it is faster, and discuss applications where it is used in modern research.

If you liked that video, hit that like button. If you wanna stick around, hit that subscribe button. If you really wanna stick around, hit that bell icon next to the subscribe button to be notified of my uploads immediately.

Convolution Neural Networks: https://www.youtube.com/watch?v=m8pOnJxOcqY

REFERENCES

Xception (main paper): https://arxiv.org/pdf/1610.02357.pdf
Mobile Nets (Efficient CNN for mobile vision applications) : https://arxiv.org/pdf/1704.04861.pdf
One model Learns all: https://arxiv.org/pdf/1706.05137v1.pdf
Music at : https://www.bensound.com/royalty-free-music/track/tenderness

Видео Depthwise Separable Convolution - A FASTER CONVOLUTION! канала CodeEmporium