arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2305.17852cs.CV
收藏

用于低延迟事件处理的分层神经记忆网络

Hierarchical Neural Memory Network for Low Latency Event Processing

National Institute of Advanced Industrial Science and Technology (AIST)(国立先进工业科学技术研究所);Simon Fraser University(西蒙弗雷泽大学)

Ryuhei Hamaguchi, Yasutaka Furukawa, Masaki Onishi, Ken Sakurada

AI总结:

本文提出分层神经记忆网络,通过多速率潜在记忆构建时间层次结构,实现基于事件密集预测的低延迟处理,并在准确性和延迟上优于现有方法。

摘要:

本文提出了一种用于基于事件的密集预测任务的低延迟神经网络架构。传统架构以固定速率编码整个场景内容,而不考虑其时间特性。相反,所提出的网络根据内容的运动速度,在适当的时间尺度上对内容进行编码。我们通过使用以不同速率运行的堆叠潜在记忆构建时间层次结构来实现这一点。给定低延迟事件流,多级记忆通过将信息从快速记忆模块传播到慢速记忆模块,逐步提取动态到静态的场景内容。该架构不仅减少了传统架构的冗余,还利用了长期依赖关系。此外,基于注意力的时间事件表示能高效地将稀疏事件流编码到记忆单元中。我们在三个基于事件的密集预测任务上进行了广泛评估,所提出的方法在准确性和延迟方面均优于现有方法,同时展示了有效的事件与图像融合能力。代码可在 https://hamarh.github.io/hmnet/ 获取。

英文摘要:

This paper proposes a low latency neural network architecture for event-based dense prediction tasks. Conventional architectures encode entire scene contents at a fixed rate regardless of their temporal characteristics. Instead, the proposed network encodes contents at a proper temporal scale depending on its movement speed. We achieve this by constructing temporal hierarchy using stacked latent memories that operate at different rates. Given low latency event steams, the multi-level memories gradually extract dynamic to static scene contents by propagating information from the fast to the slow memory modules. The architecture not only reduces the redundancy of conventional architectures but also exploits long-term dependencies. Furthermore, an attention-based event representation efficiently encodes sparse event streams into the memory cells. We conduct extensive evaluations on three event-based dense prediction tasks, where the proposed approach outperforms the existing methods on accuracy and latency, while demonstrating effective event and image fusion capabilities. The code is available at https://hamarh.github.io/hmnet/

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑