Three Steps to Multimodal Trajectory Prediction: Modality Clustering, Classification and Synthesis
This work addresses the need for accurate multimodal predictions in trajectory forecasting, which is crucial for applications like autonomous driving, and presents a novel framework that significantly outperforms prior state-of-the-art.
The paper tackles the multimodal trajectory prediction problem by proposing a new framework that divides it into three steps: modality clustering, classification, and synthesis, addressing weaknesses in existing methods. It achieves a 19.2% improvement in ADE and 20.8% in FDE on the ETH/UCY dataset without using social or map information.
Multimodal prediction results are essential for trajectory prediction task as there is no single correct answer for the future. Previous frameworks can be divided into three categories: regression, generation and classification frameworks. However, these frameworks have weaknesses in different aspects so that they cannot model the multimodal prediction task comprehensively. In this paper, we present a novel insight along with a brand-new prediction framework by formulating multimodal prediction into three steps: modality clustering, classification and synthesis, and address the shortcomings of earlier frameworks. Exhaustive experiments on popular benchmarks have demonstrated that our proposed method surpasses state-of-the-art works even without introducing social and map information. Specifically, we achieve 19.2% and 20.8% improvement on ADE and FDE respectively on ETH/UCY dataset. Our code will be made publicly availabe.