CERBERUS: Simple and Effective All-In-One Automotive Perception Model with Multi Task Learning
This work addresses the problem of high computational demands for autonomous driving perception on resource-constrained in-vehicle systems, representing an incremental improvement through multitask learning.
The authors tackled the computational inefficiency of running multiple deep learning models for automotive perception tasks on embedded platforms by developing CERBERUS, a lightweight multitask model that performs multiple tasks with a single inference, achieving competitive performance with reduced computational cost.
Perceiving the surrounding environment is essential for enabling autonomous or assisted driving functionalities. Common tasks in this domain include detecting road users, as well as determining lane boundaries and classifying driving conditions. Over the last few years, a large variety of powerful Deep Learning models have been proposed to address individual tasks of camera-based automotive perception with astonishing performances. However, the limited capabilities of in-vehicle embedded computing platforms cannot cope with the computational effort required to run a heavy model for each individual task. In this work, we present CERBERUS (CEnteR Based End-to-end peRception Using a Single model), a lightweight model that leverages a multitask-learning approach to enable the execution of multiple perception tasks at the cost of a single inference. The code will be made publicly available at https://github.com/cscribano/CERBERUS