Two-stream convolutional networks for end-to-end learning of self-driving cars
This work addresses the challenge of learning spatiotemporal features for autonomous driving, though it appears incremental as it builds on established two-stream networks.
The paper tackled the problem of end-to-end learning for self-driving cars by extending Two-Stream Convolutional Networks to incorporate temporal cues, resulting in a 30% improvement in prediction accuracy and stability compared to existing regression methods on the Comma.ai dataset.
We propose a methodology to extend the concept of Two-Stream Convolutional Networks to perform end-to-end learning for self-driving cars with temporal cues. The system has the ability to learn spatiotemporal features by simultaneously mapping raw images and pre-calculated optical flows directly to steering commands. Although optical flows encode temporal-rich information, we found that 2D-CNNs are prone to capturing features only as spatial representations. We show how the use of Multitask Learning favors the learning of temporal features via inductive transfer from a shared spatiotemporal representation. Preliminary results demonstrate a competitive improvement of 30% in prediction accuracy and stability compared to widely used regression methods trained on the Comma.ai dataset.