1: \begin{abstract}

2: 	We present the first distributed optimization algorithm with \emph{lazy communication} for collaborative geometric estimation, the backbone of modern collaborative simultaneous localization and mapping (SLAM) and structure-from-motion (SfM) applications. 

3: 	Our method allows agents to cooperatively reconstruct a shared geometric model on a central server by fusing individual observations, 

4: 	but \emph{without} the need to transmit potentially sensitive information about the agents themselves (such as their locations).

5: 	Furthermore, to alleviate the burden of communication during iterative optimization, we design a set of \emph{communication triggering conditions} that enable agents to selectively upload \revise{a targeted subset of} local information that \revise{is} useful to global optimization. 

6: 	Our approach thus achieves significant communication reduction with minimal impact on optimization performance.

7: 	As our main theoretical contribution, we prove that our method converges to first-order critical points with a \revise{global} sublinear convergence rate.

8: 	Numerical evaluations on bundle adjustment problems from collaborative SLAM and SfM datasets show that our method performs competitively against existing distributed techniques, while achieving up to 78\% total communication reduction.

9: \end{abstract}

10: