Methods for source detection in high noise environments are crucial for single-photon emission computed tomography (SPECT) medical imaging and especially for homeland security applications, which is our main interest. In the latter case, one deals with detecting the presence of low emission nuclear sources with significant background noise (with Signal To Noise Ratio ($SNR$) $1\%$ or less). Direction sensitive detectors are needed to achieve this goal. Collimation, used for that purpose in standard $\gamma$-cameras, is not an option. Instead, Compton cameras can be utilized. Backprojection methods suggested before enable detection in the presence of a random uniform background. In most practical applications, however, cargo packing in shipping containers and trucks creates regions of strong absorption, while leaving streaming gaps open. In such cases the background will not be uniform, which renders backprojection methods ineffective. A deep neural network is implemented for the source detection in 2D, which exhibits higher sensitivity and specificity than the backprojection techniques in a low scattering case and works well when presence of cargo makes the latter fail.