Implementation of CSRNet Crowd Counting Project for Udacity Project Showcase
In this article i would like to discuss about the CSRNet & i followed the blog written by Pulkit Sharma in Analytics Vidhya.
The Paper is CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes, https://arxiv.org/abs/1802.10062
This paper mainly focuses on Dilated Convolutions, below is some theory so you can get some idea.
Dilated Convolution
Standard Convolution (Left), Dilated Convolution (Right)
The left one is the standard convolution. The right one is the dilated convolution. We can see that at the summation, it is s+lt=p that we will skip some points during convolution.
When l=1, it is standard convolution.
When l>1, it is dilated convolution.
Standard Convolution (l=1)
Dilated Convolution (l=2)
The above illustrate an example of dilated convolution when l=2. We can see that the receptive field is larger compared with the standard one.
l=1 (left), l=2 (Middle), l=4 (Right)
The above figure shows more examples about the receptive field.
So, by using Dilated Convolutions, you can get the wider Receptive field than normal.
I followed this article please have a look: https://www.analyticsvidhya.com/blog/2019/02/building-crowd-counting-model-python/ & you can try to implement the same.
Output looks like this:
References:
