Geo-Motion Data Association for Reliable Vehicle Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle tracking systems rely heavily on priors that may be incorrect, leading to unreliable predictions, and often ignore geometry and motion information, resulting in unstable and inaccurate object association.
Innovation Solution
A system that uses geo-motion embeddings and appearance embeddings processed by neural networks to associate new object measurements with maintained tracks, leveraging both geometric and motion information, and appearance data without relying on prior distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing autonomous vehicle tracking systems rely heavily on priors for object association, then the system can make predictions, but the predictions become unreliable when priors are incorrect
Solution Approach 1:
The patent extracts and removes the dependency on prior distributions from the tracking system. Instead of using probabilistic data association with priors, the system directly processes geo-motion embeddings and appearance embeddings through neural networks to generate association predictions, eliminating the source of unreliability when priors are incorrect
Solution Approach 2:
The patent substitutes the mechanical/probabilistic system of prior-based data association with a neural network-based system. The neural networks process embeddings and generate predictions without relying on predefined probability distributions, replacing the old mechanism with a learning-based approach that adapts to actual data patterns
2Measurement precision
If tracking systems ignore geometry and motion information, then the system simplifies processing, but object association becomes unstable and inaccurate
Solution Approach 1:
The patent merges geometry information (from geo-motion embeddings containing position, size, shape) and motion information (velocity, acceleration) with appearance information. This combination is processed jointly through neural networks to produce more accurate and stable object association predictions than using appearance alone
Solution Approach 2:
The patent performs preliminary processing to extract geo-motion embeddings and appearance embeddings before feeding them to the association neural networks. This pre-processing organizes geometry, motion, and appearance data into structured embeddings, making the subsequent association processing more efficient and accurate
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating a new measurement of an object surrounding a vehicle with a maintained track. One of the methods includes receiving an object track for a particular object, receiving a new measurement characterizing a new object at a new time step, and determining whether the new object is the same as the particular object, comprising: generating a representation of the new object at the new and preceding time steps; generating a representation of the particular object at the new and preceding time steps; processing a first network input comprising the representations using a first neural network to generate an embedding of the first network input; and processing the embedding of the first network input using a second neural network to generate a predicted likelihood that the new object and the particular object are the same.