Consistency-based Semi-supervised Learning for Object detection

Authors: Jisoo Jeong, Seungeui Lee, Jeesoo Kim, Nojun Kwak

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We have evaluated the proposed CSD both in single-stage and two-stage detectors and the results show the effectiveness of our method.
Researcher Affiliation Academia Jisoo Jeong , Seungeui Lee , Jeesoo Kim & Nojun Kwak Department of Transdisciplinary Studies Graduate School of Convergence Science and Technology Seoul National University Seoul, Korea
Pseudocode No The paper describes the method using textual descriptions, equations, and diagrams, but does not include any structured pseudocode or algorithm blocks.
Open Source Code Yes 2https://github.com/soo89/CSD-SSD 3https://github.com/soo89/CSD-RFCN
Open Datasets Yes In our experiments, we have utilized the PASCAL VOC [22] and MSCOCO [23] datasets which are the most popular datasets in object detection.
Dataset Splits Yes PASCAL VOC 2007 and 2012 datasets consist of 5,011 and 11,540 trainval (train and validation) images respectively. In this paper, PASCAL VOC2007 trainval is used as the labeled data and PASCAL VOC2012 trainval and MSCOCO are utilized as the unlabeled one.
Hardware Specification No The paper does not provide specific details about the hardware used for experiments, such as GPU models or CPU types.
Software Dependencies No The paper mentions that the codes are based on PyTorch, and uses third-party codes for SSD and R-FCN, but it does not specify version numbers for PyTorch or any other software dependencies.
Experiment Setup No The paper states that 'all the parameter settings and training details are presented in the supplementary material' and 'Details on learning scheduling are in the supplementary material,' but these specific details are not provided in the main text.