SECURE: Stable Early Collision Understanding via Robust Embeddings in Autonomous Driving

TL;DR

SECURE is a framework that enhances the robustness and stability of accident anticipation models in autonomous driving by enforcing consistency in predictions and features, leading to improved reliability.

Contribution

The paper introduces SECURE, a novel training methodology that improves model robustness against perturbations while also boosting accuracy on clean data.

Findings

SECURE significantly improves robustness against input perturbations.

The approach achieves state-of-the-art results on DAD and CCD datasets.

Model stability and performance are both enhanced by the proposed method.

Abstract

While deep learning has significantly advanced accident anticipation, the robustness of these safety-critical systems against real-world perturbations remains a major challenge. We reveal that state-of-the-art models like CRASH, despite their high performance, exhibit significant instability in predictions and latent representations when faced with minor input perturbations, posing serious reliability risks. To address this, we introduce SECURE - Stable Early Collision Understanding Robust Embeddings, a framework that formally defines and enforces model robustness. SECURE is founded on four key attributes: consistency and stability in both prediction space and latent feature space. We propose a principled training methodology that fine-tunes a baseline model using a multi-objective loss, which minimizes divergence from a reference model and penalizes sensitivity to adversarial perturbations. Experiments on DAD and CCD datasets demonstrate that our approach not only significantly enhances robustness against various perturbations but also improves performance on clean data, achieving new state-of-the-art results.