SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos

TL;DR

SAVi++ is an advanced object-centric model that learns to segment and track objects in complex real-world videos using depth signals, without explicit segmentation supervision, and can leverage LiDAR data for improved performance.

Contribution

The paper introduces SAVi++, a novel end-to-end model that uses depth signals to enable scalable object segmentation and tracking in real-world videos without supervision.

Findings

Successfully segments complex scenes with moving cameras and diverse objects.

Learns from naturalistic backgrounds without explicit segmentation labels.

Utilizes LiDAR data to enhance real-world object tracking.

Abstract

The visual world can be parsimoniously characterized in terms of distinct entities with sparse interactions. Discovering this compositional structure in dynamic visual scenes has proven challenging for end-to-end computer vision approaches unless explicit instance-level supervision is provided. Slot-based models leveraging motion cues have recently shown great promise in learning to represent, segment, and track objects without direct supervision, but they still fail to scale to complex real-world multi-object videos. In an effort to bridge this gap, we take inspiration from human development and hypothesize that information about scene geometry in the form of depth signals can facilitate object-centric learning. We introduce SAVi++, an object-centric video model which is trained to predict depth signals from a slot-based video representation. By further leveraging best practices for model scaling, we are able to train SAVi++ to segment complex dynamic scenes recorded with moving cameras, containing both static and moving objects of diverse appearance on naturalistic backgrounds, without the need for segmentation supervision. Finally, we demonstrate that by using sparse depth signals obtained from LiDAR, SAVi++ is able to learn emergent object segmentation and tracking from videos in the real-world Waymo Open dataset.