Depth Estimation using DNN Architecture and Vision-Based Transformers

Depth Estimation is the process of estimating depth objects with a 2D image as an input. The importance that it provides some critical information about 3D structure scene given sequence images. This helpful in various applications like robotics, virtual reality, autonomous driving, medical imaging, and so on forth. Our main objective here to make environment more perceivable by vehicles. Unlike traditional approaches which try extract from input image, we apply instance segmentation then use segmented map generator. In this paper, fully connected neural network dense prediction transformers. As mentioned above, after for Transformer backbone producing high-resolution experimentation results have shown many improvements related details sharpness compared estimation techniques. architecture applied tested KITTI data set.