This document discusses object detection under rainy conditions and techniques to address challenges with training and detecting objects from images distorted by rain. It summarizes popular object detection architectures like Faster R-CNN and YOLO. It also reviews state-of-the-art image de-raining algorithms like Deep Detail Network, Attentive GAN, and Progressive Image Deraining Network that aim to remove rain distortions while maintaining image details. Finally, it mentions unsupervised image translation and domain adaptation as alternative training approaches for deep learning object detection under adverse weather conditions.