The paper titled "Towards Learning Monocular 3D Object Localization Using the Physical Laws of Motion" by Daniel Kienzle, Julian Lorenz, Katja Ludwig and Rainer Lienhart was accepted to the International Conference on 3D Vision (3DV) 2024. The paper describes a new method for localizing objects in 3D without the need for 3D ground truth. Instead, the method uses knowledge of physical laws to learn the task.

 

See https://kiedani.github.io/3DV2024/ for additional information on the paper.

 

We present a novel method for precise 3D object localization in single images from a single calibrated camera using only 2D labels. No expensive 3D labels are needed. Thus, instead of using 3D labels, our model is trained with easy-to-annotate 2D labels along with the physical knowledge of the object's motion. Given this information, the model can infer the latent third dimension, even though it has never seen this information during training. Our method is evaluated on both synthetic and real-world datasets, and we are able to achieve a mean distance error of just 6 cm in our experiments on real data. The results indicate the method's potential as a step towards learning 3D object location estimation, where collecting 3D data for training is not feasible.