HCNAF: Hyper-Conditioned Neural Autoregressive Flow and its Application for Probabilistic Occupancy Map Forecasting
This work addresses the need for accurate probabilistic forecasting in autonomous driving, representing an incremental improvement in flow-based models for high-dimensional prediction tasks.
The authors tackled the problem of modeling complex conditional probability distributions by introducing Hyper-Conditioned Neural Autoregressive Flow (HCNAF), which achieved state-of-the-art performance on a public self-driving dataset for probabilistic occupancy map forecasting.
We introduce Hyper-Conditioned Neural Autoregressive Flow (HCNAF); a powerful universal distribution approximator designed to model arbitrarily complex conditional probability density functions. HCNAF consists of a neural-net based conditional autoregressive flow (AF) and a hyper-network that can take large conditions in non-autoregressive fashion and outputs the network parameters of the AF. Like other flow models, HCNAF performs exact likelihood inference. We conduct a number of density estimation tasks on toy experiments and MNIST to demonstrate the effectiveness and attributes of HCNAF, including its generalization capability over unseen conditions and expressivity. Finally, we show that HCNAF scales up to complex high-dimensional prediction problems of the magnitude of self-driving and that HCNAF yields a state-of-the-art performance in a public self-driving dataset.