Occluded Object Tracking via Frame State Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object tracking systems struggle to accurately track objects that become occluded for multiple frames, leading to fragmented tracking and increased risk of collisions in autonomous navigation.
Innovation Solution
An object tracking model that encodes locations of visible objects in a sequence of frames, generates a representation of the environment's state, and predicts the location of occluded objects by comparing object centers across prior frames, allowing for robust tracking even when objects are partially or fully occluded.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional object tracking systems use standard detection and tracking methods, then visible objects can be detected and tracked, but tracking becomes fragmented when objects are occluded for multiple frames
Solution Approach 1:
The system performs preliminary encoding of object locations and appearances during visible frames, storing this information in advance. When occlusion occurs, the pre-encoded information from previous frames is retrieved and used to maintain continuous tracking, preventing fragmentation of the tracking trajectory.
Solution Approach 2:
The system introduces an intermediary representation layer that encodes object states beyond direct visual detection. This intermediary representation maintains object identity and location information through occlusion periods by comparing encoded representations across frames, acting as a mediator between visible and occluded states.
2Measurement precision
If the system encodes and compares object centers across multiple prior frames, then occluded object locations can be predicted, but computational complexity increases
Solution Approach 1:
The system extracts only the essential object center location information from full object representations. By focusing on comparing object centers rather than entire object models, the system achieves accurate occluded object prediction while significantly reducing computational complexity and memory requirements.
Solution Approach 2:
The system changes the parameter representation from complete object models to simplified object center coordinates. This parameter transformation enables efficient comparison and prediction operations across multiple frames while maintaining the necessary precision for tracking occluded objects.
Data Source
AI summary
A method for tracking an object performed by an object tracking system includes encoding locations of visible objects in an environment captured in a current frame of a sequence of frames. The method also includes generating a representation of a current state of the environment based on an aggregation of the encoded locations and an encoded location of each object visible in one or more frames of the sequence of frames occurring prior to the current frame. The method further includes predicting a location of an object occluded in the current frame based on a comparison of object centers decoded from the representation of the current state to object centers saved from each prior representation associated with a different respective frame of the sequence of frames occurring prior to the current frame. The method still further includes adjusting a behavior of an autonomous agent in response to identifying the location of the occluded object.