Global Clipper: Enhancing Safety and Reliability of Transformer-based Object Detection Models

TL;DR

This paper introduces the Global Clipper and Global Hybrid Clipper strategies to improve the safety and reliability of transformer-based object detection models against soft errors, achieving near-zero faulty inferences.

Contribution

It presents novel mitigation techniques specifically designed for transformer models, addressing their unique vulnerabilities and outperforming traditional CNN solutions.

Findings

Significantly reduces faulty inferences to around 0%

Extensive testing across 64 scenarios with 3.3 million inferences

Demonstrates robustness improvements in transformer models like DINO-DETR and Lite-DETR

Abstract

As transformer-based object detection models progress, their impact in critical sectors like autonomous vehicles and aviation is expected to grow. Soft errors causing bit flips during inference have significantly impacted DNN performance, altering predictions. Traditional range restriction solutions for CNNs fall short for transformers. This study introduces the Global Clipper and Global Hybrid Clipper, effective mitigation strategies specifically designed for transformer-based models. It significantly enhances their resilience to soft errors and reduces faulty inferences to ~ 0\%. We also detail extensive testing across over 64 scenarios involving two transformer models (DINO-DETR and Lite-DETR) and two CNN models (YOLOv3 and SSD) using three datasets, totalling approximately 3.3 million inferences, to assess model robustness comprehensively. Moreover, the paper explores unique aspects of attention blocks in transformers and their operational differences from CNNs.