Cannot See the Forest for the Trees: Aggregating Multiple Viewpoints to Better Classify Objects in Videos

TL;DR

This paper introduces a set classifier that aggregates multiple viewpoints within tracklets to improve object classification in videos, especially under long-tailed distributions, achieving state-of-the-art results.

Contribution

The paper proposes a novel set classifier and tracklet augmentation method that enhance long-tailed object tracking performance by leveraging multiple viewpoints and handling sparse annotations.

Findings

Achieved new state-of-the-art TrackAP_50 scores on TAO dataset.

Significantly improved classification accuracy in long-tailed video object tracking.

Demonstrated plug-and-play compatibility with existing trackers.

Abstract

Recently, both long-tailed recognition and object tracking have made great advances individually. TAO benchmark presented a mixture of the two, long-tailed object tracking, in order to further reflect the aspect of the real-world. To date, existing solutions have adopted detectors showing robustness in long-tailed distributions, which derive per-frame results. Then, they used tracking algorithms that combine the temporally independent detections to finalize tracklets. However, as the approaches did not take temporal changes in scenes into account, inconsistent classification results in videos led to low overall performance. In this paper, we present a set classifier that improves accuracy of classifying tracklets by aggregating information from multiple viewpoints contained in a tracklet. To cope with sparse annotations in videos, we further propose augmentation of tracklets that can maximize data efficiency. The set classifier is plug-and-playable to existing object trackers, and highly improves the performance of long-tailed object tracking. By simply attaching our method to QDTrack on top of ResNet-101, we achieve the new state-of-the-art, 19.9% and 15.7% TrackAP_50 on TAO validation and test sets, respectively.