While initial approaches to Structure-from-Motion (SfM) revolved around both global and incremental methods, most recent applications rely on incremental systems to estimate camera poses due to their superior robustness. Though there has been tremendous progress in SfM `front-ends' powered by deep models learned from data, the state-of-the-art (incremental) SfM pipelines still rely on classical SIFT features, developed in 2004. In this work, we investigate whether leveraging the developments in feature extraction and matching helps global SfM perform on par with the SOTA incremental SfM approach (COLMAP). To do so, we design a modular SfM framework that allows us to easily combine developments in different stages of the SfM pipeline. Our experiments show that while developments in deep-learning based two-view correspondence estimation do translate to improvements in point density for scenes reconstructed with global SfM, none of them outperform SIFT when comparing with incremental SfM results on a range of datasets. Our SfM system is designed from the ground up to leverage distributed computation, enabling us to parallelize computation on multiple machines and scale to large scenes.

Ambient.ai is an AI company headquartered in Palo Alto on a mission to prevent as many security incidents as possible. Our breakthrough technology combines cutting-edge deep learning with a contextual knowledge model to achieve human-like perception ability. Ambient's flagship product has been deployed by multiple Fortune 100 companies to solve a mission-critical problem in a way that has never been possible. The company was founded in 2017 by Shikhar Shrestha and Vikesh Khanna who are experts in artificial intelligence from Stanford University who previously built iconic products at Apple, Google, Microsoft, and Dropbox. We are a Series-B company backed by Andreessen Horowitz (a16z), SV Angel, YCombinator, and visionary angels like Jyoti Bansal, Mark Leslie, and Elad Gil.

This function takes the task type, data set, parameters and GPU node which users enter to launch a training task in the GPU cluster for users to monitor and manage. The panel for this function is as below. Supported task types include Image Classification and Depth Estimation. Select a data set that you have uploaded to the server in the designated format, a GPU node and training parameters before clicking Start Training! This function allows users to monitor running tasks, model convergence and machine status.