Diverse Complexity Measures for Dataset Curation in Self-driving

by   Abbas Sadat, et al.

Modern self-driving autonomy systems heavily rely on deep learning. As a consequence, their performance is influenced significantly by the quality and richness of the training data. Data collecting platforms can generate many hours of raw data in a daily basis, however, it is not feasible to label everything. It is thus of key importance to have a mechanism to identify "what to label". Active learning approaches identify examples to label, but their interestingness is tied to a fixed model performing a particular task. These assumptions are not valid in self-driving, where we have to solve a diverse set of tasks (i.e., perception, and motion forecasting) and our models evolve over time frequently. In this paper we introduce a novel approach and propose a new data selection method that exploits a diverse set of criteria that quantize interestingness of traffic scenes. Our experiments on a wide range of tasks and models show that the proposed curation pipeline is able to select datasets that lead to better generalization and higher performance.


page 1

page 2

page 3

page 4


Just Label What You Need: Fine-Grained Active Selection for Perception and Prediction through Partially Labeled Scenes

Self-driving vehicles must perceive and predict the future positions of ...

LookOut: Diverse Multi-Future Prediction and Planning for Self-Driving

Self-driving vehicles need to anticipate a diverse set of future traffic...

One Thousand and One Hours: Self-driving Motion Prediction Dataset

We present the largest self-driving dataset for motion prediction to dat...

Data-Efficient Learning via Minimizing Hyperspherical Energy

Deep learning on large-scale data is dominant nowadays. The unprecedente...

Recovering and Simulating Pedestrians in the Wild

Sensor simulation is a key component for testing the performance of self...

Rethinking Self-driving: Multi-task Knowledge for Better Generalization and Accident Explanation Ability

Current end-to-end deep learning driving models have two problems: (1) P...