Hierarchical Instance Tracking to Balance Privacy Preservation with Accessible Information
This addresses the need for structured visual tracking in applications like robotics or surveillance, but it is incremental as it builds on existing tracking methods with a new dataset and task definition.
The authors introduced hierarchical instance tracking, a new task for tracking objects and parts while preserving their hierarchical relationships, and created the first benchmark dataset with 2,765 entities across 552 videos and 40 categories, finding it challenging for evaluated models.
We propose a novel task, hierarchical instance tracking, which entails tracking all instances of predefined categories of objects and parts, while maintaining their hierarchical relationships. We introduce the first benchmark dataset supporting this task, consisting of 2,765 unique entities that are tracked in 552 videos and belong to 40 categories (across objects and parts). Evaluation of seven variants of four models tailored to our novel task reveals the new dataset is challenging. Our dataset is available at https://vizwiz.org/tasks-and-datasets/hierarchical-instance-tracking/