ICML2024科研海报设计
High-Performance Temporal Reversible Spiking Neural Networks with O(L) Training Memory and O(1) Inference Cost
一张来自 ICML 2024 的真实会议海报,作为科研海报设计参考展示。研读作者如何排布标题、图表和结果,再设计你自己的海报。
读者能从这张海报里答出这些吗?
设计得当的海报能让路过的人几秒内抓住这些要点——可作为你自己科研海报设计的检查清单。
- →What is the primary advantage of Spiking Neural Networks (SNNs) over Artificial Neural Networks (ANNs)?
- →What problem does the study aim to address in the context of SNNs?
- →What is the primary goal of the T-RevSNN architecture proposed in the study?
- →What is a key novelty of the T-RevSNN architecture?
- →How does the T-RevSNN architecture achieve O(1) inference energy cost?
