Why not densenet?

[sanchit2843]

I dont know whether this is the right place for this question and I am new to video classification. Most of the recent implementations are still using resnet based 3-d architecture, why the dense net is not used yet?