Self-Consistency for LLM-Based Motion Trajectory Generation and Verification

TL;DR

This paper introduces a self-consistency approach for improving LLM-based motion trajectory generation and verification by modeling shape families with geometric transformations, leading to significant accuracy and precision gains.

Contribution

It adapts self-consistency from language reasoning to visual motion trajectories, proposing an algorithm to recover shape families and enhance generation and verification accuracy.

Findings

Improves LLM trajectory generation accuracy by 4-6%.

Extends self-consistency to motion verification with 11% precision gains.

Provides a new algorithm for modeling shape families with geometric transformations.

Abstract

Self-consistency has proven to be an effective technique for improving LLM performance on natural language reasoning tasks in a lightweight, unsupervised manner. In this work, we study how to adapt self-consistency to visual domains. Specifically, we consider the generation and verification of LLM-produced motion graphics trajectories. Given a prompt (e.g., "Move the circle in a spiral path"), we first sample diverse motion trajectories from an LLM, and then identify groups of consistent trajectories via clustering. Our key insight is to model the family of shapes associated with a prompt as a prototype trajectory paired with a group of geometric transformations (e.g., rigid, similarity, and affine). Two trajectories can then be considered consistent if one can be transformed into the other under the warps allowable by the transformation group. We propose an algorithm that automatically recovers a shape family, using hierarchical relationships between a set of candidate transformation groups. Our approach improves the accuracy of LLM-based trajectory generation by 4-6%. We further extend our method to support verification, observing 11% precision gains over VLM baselines. Our code and dataset are available at https://majiaju.io/trajectory-self-consistency .