Semi-supervised Class-Agnostic Motion Prediction with Pseudo Label Regeneration and BEVMix

Authors: Kewei Wang, Yizheng Wu, Zhiyu Pan, Xingyi Li, Ke Xian, Zhe Wang, Zhiguo Cao, Guosheng Lin

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments conducted on nu Scenes demonstrate that our SSL method can surpass the self-supervised approach by a large margin by utilizing only a tiny fraction of labeled data. Furthermore, our method exhibits comparable performance to weakly and some fully supervised methods.
Researcher Affiliation Collaboration 1 Key Laboratory of Image Processing and Intelligent Control, Ministry of Education School of Artificial Intelligence and Automation, Huazhong University of Science and Technology 2 S-Lab, Nanyang Technological University 3 Sense Time Research
Pseudocode Yes Algorithm 1: Pseudo Label Re-generation
Open Source Code Yes Code will be available at https://github.com/kwwcv/SSMP.
Open Datasets Yes Following previous work... nu Scenes (Caesar et al. 2019), which contains 850 scenes with annotations.
Dataset Splits Yes For fair comparisons, we follow Motion Net to use 500 scenes for training, 100 scenes for validation, and 250 scenes for testing.
Hardware Specification Yes We implement our model in Pytorch (Paszke et al. 2019) with a single A6000 GPU.
Software Dependencies No The paper mentions "Pytorch" and cites its authors (Paszke et al. 2019), but it does not specify a version number for Pytorch. It also mentions "Adam" as an optimizer, which is an algorithm, not a software dependency with a version.
Experiment Setup Yes The parameters K, µ, β, γ, θc, θw, and α are set to 5, 1, 10, 0.6, 3, 5, and 0.999, respectively.