Guardado en:
Detalles Bibliográficos
Autores principales: Li, Yuxin, Li, Yiheng, Yang, Xulei, Yu, Mengying, Huang, Zihang, Wu, Xiaojun, Yeo, Chai Kiat
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:https://arxiv.org/abs/2410.07268
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914969118834688
author Li, Yuxin
Li, Yiheng
Yang, Xulei
Yu, Mengying
Huang, Zihang
Wu, Xiaojun
Yeo, Chai Kiat
author_facet Li, Yuxin
Li, Yiheng
Yang, Xulei
Yu, Mengying
Huang, Zihang
Wu, Xiaojun
Yeo, Chai Kiat
contents In the landscape of autonomous driving, Bird's-Eye-View (BEV) representation has recently garnered substantial academic attention, serving as a transformative framework for the fusion of multi-modal sensor inputs. This BEV paradigm effectively shifts the sensor fusion challenge from a rule-based methodology to a data-centric approach, thereby facilitating more nuanced feature extraction from an array of heterogeneous sensors. Notwithstanding its evident merits, the computational overhead associated with BEV-based techniques often mandates high-capacity hardware infrastructures, thus posing challenges for practical, real-world implementations. To mitigate this limitation, we introduce a novel content-aware multi-modal joint input pruning technique. Our method leverages BEV as a shared anchor to algorithmically identify and eliminate non-essential sensor regions prior to their introduction into the perception model's backbone. We validatethe efficacy of our approach through extensive experiments on the NuScenes dataset, demonstrating substantial computational efficiency without sacrificing perception accuracy. To the best of our knowledge, this work represents the first attempt to alleviate the computational burden from the input pruning point.
format Preprint
id arxiv_https___arxiv_org_abs_2410_07268
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Learning Content-Aware Multi-Modal Joint Input Pruning via Bird's-Eye-View Representation
Li, Yuxin
Li, Yiheng
Yang, Xulei
Yu, Mengying
Huang, Zihang
Wu, Xiaojun
Yeo, Chai Kiat
Computer Vision and Pattern Recognition
Artificial Intelligence
In the landscape of autonomous driving, Bird's-Eye-View (BEV) representation has recently garnered substantial academic attention, serving as a transformative framework for the fusion of multi-modal sensor inputs. This BEV paradigm effectively shifts the sensor fusion challenge from a rule-based methodology to a data-centric approach, thereby facilitating more nuanced feature extraction from an array of heterogeneous sensors. Notwithstanding its evident merits, the computational overhead associated with BEV-based techniques often mandates high-capacity hardware infrastructures, thus posing challenges for practical, real-world implementations. To mitigate this limitation, we introduce a novel content-aware multi-modal joint input pruning technique. Our method leverages BEV as a shared anchor to algorithmically identify and eliminate non-essential sensor regions prior to their introduction into the perception model's backbone. We validatethe efficacy of our approach through extensive experiments on the NuScenes dataset, demonstrating substantial computational efficiency without sacrificing perception accuracy. To the best of our knowledge, this work represents the first attempt to alleviate the computational burden from the input pruning point.
title Learning Content-Aware Multi-Modal Joint Input Pruning via Bird's-Eye-View Representation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2410.07268