Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2406.03229 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916317111517184 |
|---|---|
| author | Sha, Qutub Syed Paulitsch, Michael Pattabiraman, Karthik Hagn, Korbinian Oboril, Fabian Buerkle, Cornelius Scholl, Kay-Ulrich Hinz, Gereon Knoll, Alois |
| author_facet | Sha, Qutub Syed Paulitsch, Michael Pattabiraman, Karthik Hagn, Korbinian Oboril, Fabian Buerkle, Cornelius Scholl, Kay-Ulrich Hinz, Gereon Knoll, Alois |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_03229 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Global Clipper: Enhancing Safety and Reliability of Transformer-based Object Detection Models Sha, Qutub Syed Paulitsch, Michael Pattabiraman, Karthik Hagn, Korbinian Oboril, Fabian Buerkle, Cornelius Scholl, Kay-Ulrich Hinz, Gereon Knoll, Alois Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning 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. |
| title | Global Clipper: Enhancing Safety and Reliability of Transformer-based Object Detection Models |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2406.03229 |