Spiking PointNet: Spiking Neural Networks for Point Clouds

Authors: Dayong Ren, Zhe Ma, Yuanpei Chen, Weihang Peng, Xiaode Liu, Yuhan Zhang, Yufei Guo

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct various experiments on Model Net10, Model Net40 to demonstrate the effectiveness of Spiking Point Net.
Researcher Affiliation Collaboration Dayong Ren, Zhe Ma, Yuanpei Chen, Weihang Peng, Xiaode Liu, Yuhan Zhang, Yufei Guo Intelligent Science & Technology Academy of CASIC, China Scientific Research Laboratory of Aerospace Intelligent Systems and Technology, China rdyedu@gmail.com, yfguo_bit@126.com, yfguo@pku.edu.cn
Pseudocode No The paper describes its methods verbally and with equations but does not include any explicitly labeled pseudocode or algorithm blocks.
Open Source Code Yes Our code is open-sourced at Spiking-Point Net.
Open Datasets Yes We conduct extensive experiments on Model Net10 and Model Net40 [50] to demonstrate the superior performance of our method. Model Net10 and Model Net40 are two widely recognized public datasets used for 3D object classification, curated and maintained by a research team at Princeton University.
Dataset Splits No The paper mentions training and inferring phases, but does not explicitly provide details about training/validation/test dataset splits (e.g., percentages, sample counts, or explicit reference to predefined validation splits).
Hardware Specification No The paper discusses energy estimation using '45nm CMOS technology' and calculates energy consumption values, but it does not specify the actual hardware (e.g., specific CPU or GPU models, type of computing cluster) used to conduct the experiments.
Software Dependencies No The paper mentions general machine learning frameworks like 'Tensor Flow, Pytorch' but does not specify their version numbers or any other software dependencies with version details required for replication.
Experiment Setup Yes For all our SNN models, we set Vth as 0.5, The initial perturbations, δ, range from 0 to 0.5. [...] In the paper, we choose k as 5.