Monocular Dense Depth Map Estimation using CNN

​

The goal is to experiment using deep learning networks to estimate depth from a single 2D image i.e Monocular depth estimation. 

​

Given a single 3 channel RGB image as input, predict a dense depth map for each pixel. ​

Given a training set T = {(Ii, Di)}M, where i= 1 to M, Ii ∈ I and Di ∈ D,

the task is to learn a non-linear mapping Φ: I → D, where I is 2D 3 channel RGB Image and D is Depth map.​

​