Region-level Contrastive and Consistency Learning for Semi-Supervised Semantic Segmentation

Authors: Jianrong Zhang, Tianyi Wu, Chuanghao Ding, Hongwei Zhao, Guodong Guo

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental and evaluate our RC2L on two challenging benchmarks (PASCAL VOC 2012 and Cityscapes), outperforming the state-of-the-art.
Researcher Affiliation Collaboration Jianrong Zhang1 , Tianyi Wu2,3 , Chuanghao Ding4 , Hongwei Zhao1 , Guodong Guo2,3 1College of Computer Science and Technology, Jilin University, Changchun, China 2Institute of Deep Learning, Baidu Research, Beijing, China 3National Engineering Laboratory for Deep Learning Technology and Application, Beijing, China 4College of Software, Jilin University, Changchun, China
Pseudocode No The paper describes methods in text and diagrams but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described, nor does it include specific repository links or explicit code release statements.
Open Datasets Yes PASCAL VOC 2012. PASCAL VOC 2012 [Everingham et al., 2015] is a standard object-centric semantic segmentation dataset... Cityscapes. Cityscapes [Cordts et al., 2016] dataset is designed for urban scene understanding...
Dataset Splits Yes For both the original and augmented datasets, 1/2, 1/4, 1/8, and 1/16 training images are used as labeled data, respectively, for conducting the semi-supervised experiments. ... The finely annotated 5,000 images that follow the official split have 2,975 images for training, 500 images for evaluation, and 1,525 images for testing.
Hardware Specification Yes All the models are trained using 4 V100 GPUs. ... All the models are trained using 8 V100 GPUs.
Software Dependencies No The paper mentions software components like 'Res Net101', 'Mask Former Head', and 'ADAMW optimizer' but does not provide specific version numbers for these or other ancillary software dependencies.
Experiment Setup Yes For all experiments, we set the batch size to 16, and use ADAMW as the optimizer with an initial learning rate of 0.0001, and weight decay of 0.0001. Empirically, we set the loss weight of β1, β2, β3 and β4 to 1, 20, 4 and 4, respectively. In addition, we employ a poly learning rate policy which is multiplied by (1 iter maxiter)power with power = 0.9. For PASCAL VOC 2012 dataset, we use the crop size of 512 512, and train our model for 120K and 160K iterations for VOC Train and VOC Aug, respectively. For Cityscapes dataset, we use the crop size of 768 768, and set training iterations as 120K without using any extra training data.