Object Discovery in Videos as Foreground Motion Clustering

TL;DR

This paper introduces a novel neural network approach that clusters foreground pixel trajectories in videos to improve object discovery and segmentation accuracy, achieving state-of-the-art results on standard datasets.

Contribution

A new pixel-trajectory recurrent neural network that learns feature embeddings for foreground motion clustering in videos, enabling improved object discovery.

Findings

Achieves state-of-the-art performance on motion segmentation datasets.

Effectively links foreground object masks across video frames.

Demonstrates the effectiveness of trajectory-based clustering for object discovery.

Abstract

We consider the problem of providing dense segmentation masks for object discovery in videos. We formulate the object discovery problem as foreground motion clustering, where the goal is to cluster foreground pixels in videos into different objects. We introduce a novel pixel-trajectory recurrent neural network that learns feature embeddings of foreground pixel trajectories linked across time. By clustering the pixel trajectories using the learned feature embeddings, our method establishes correspondences between foreground object masks across video frames. To demonstrate the effectiveness of our framework for object discovery, we conduct experiments on commonly used datasets for motion segmentation, where we achieve state-of-the-art performance.