Haar Wavelet based Block Autoregressive Flows for Trajectories
This work addresses trajectory prediction for autonomous agents, offering an incremental improvement in modeling dependency structures using normalizing flows.
The paper tackles the challenge of accurately modeling multimodal trajectory distributions over long time horizons by introducing a Haar wavelet-based block autoregressive flow model, which achieves improved performance in generating diverse and accurate trajectories on real-world datasets like Stanford Drone and Intersection Drone.
Prediction of trajectories such as that of pedestrians is crucial to the performance of autonomous agents. While previous works have leveraged conditional generative models like GANs and VAEs for learning the likely future trajectories, accurately modeling the dependency structure of these multimodal distributions, particularly over long time horizons remains challenging. Normalizing flow based generative models can model complex distributions admitting exact inference. These include variants with split coupling invertible transformations that are easier to parallelize compared to their autoregressive counterparts. To this end, we introduce a novel Haar wavelet based block autoregressive model leveraging split couplings, conditioned on coarse trajectories obtained from Haar wavelet based transformations at different levels of granularity. This yields an exact inference method that models trajectories at different spatio-temporal resolutions in a hierarchical manner. We illustrate the advantages of our approach for generating diverse and accurate trajectories on two real-world datasets - Stanford Drone and Intersection Drone.