Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..

Fully Spiking Neural Networks for Unified Frame-Event Object Tracking

Authors: Jingjun Yang, Liangwei Fan, Jinpu Zhang, Xiangkai Lian, Hui Shen, Dewen Hu

NeurIPS 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments across multiple benchmarks demonstrate that the proposed framework achieves superior tracking accuracy over existing methods while significantly reducing power consumption, attaining an optimal balance between performance and efficiency.
Researcher Affiliation Academia Jingjun Yang Liangwei Fan Jinpu Zhang Xiangkai Lian Hui Shen Dewen Hu College of Intelligence Science and Technology National University of Defense Technology EMAIL
Pseudocode No The paper describes methods using mathematical formulations (e.g., equations 1-15) and block diagrams (Figure 3), but it does not contain a clearly labeled pseudocode or algorithm block.
Open Source Code Yes Codes and models: https://github.com/Noctis-A/Spike FET
Open Datasets Yes We evaluated our proposed Spike FET using three large-scale frame-event single-object tracking datasets: FE108 [8], Vis Event [6], and COESOT [7].
Dataset Splits Yes The FE108 dataset [8] ... is divided into 76 training sequences and 32 testing sequences. The Vis Event [6] dataset originally collected 820 frame-event pairs, divided into 500 training sequences and 320 testing sequences. ... refined dataset of 205 training sequences and 172 testing sequences. ... COESOT [7] ... divided into 827 training sequences and 527 testing sequences.
Hardware Specification Yes Our proposed Spike FET was implemented with Py Torch 1.12 in Python 3.8 and trained on two NVIDIA RTX 4090 GPUs.
Software Dependencies Yes Our proposed Spike FET was implemented with Py Torch 1.12 in Python 3.8 and trained on two NVIDIA RTX 4090 GPUs.
Experiment Setup Yes The model was trained using the Adam W optimizer. The optimizer has a cosine annealing scheduling over 50 epochs, where each epoch contained 60,000 image triplets. Please refer to the Appendix D for more details.