Time Cell Inspired Temporal Codebook in Spiking Neural Networks for Enhanced Image Generation

Summary:

This paper proposes a temporal codebook for spiking generative models, inspired by hippocampal time cells. The design equips an SNN-based VQ-VAE with stronger temporal representation ability, leading to better image synthesis quality and more consistent generation across diverse datasets.

Bibtex:

@misc{feng2024temporalcodebook,
  title={Time Cell Inspired Temporal Codebook in Spiking Neural Networks for Enhanced Image Generation},
  author={Linghao Feng and Dongcheng Zhao and Sicheng Shen and Yiting Dong and Guobin Shen and Yi Zeng},
  year={2024},
  eprint={2405.14474},
  archivePrefix={arXiv},
  primaryClass={cs.NE},
  url={https://arxiv.org/abs/2405.14474}
}

Recommended citation: Feng, L., Zhao, D., Shen, S., Dong, Y., Shen, G., & Zeng, Y. (2024). Time Cell Inspired Temporal Codebook in Spiking Neural Networks for Enhanced Image Generation.
Download Paper