Complete Instances Mining for Weakly Supervised Instance Segmentation

Authors: Zecheng Li, Zening Zeng, Yuqi Liang, Jin-Gang Yu

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical evaluations on the PASCAL VOC 2012 and MS COCO datasets demonstrate that our method achieves state-of-the-art performance with a notable margin.
Researcher Affiliation Academia Zecheng Li1 , Zening Zeng1 , Yuqi Liang1 and Jin-Gang Yu1,2 1South China University of Technology 2Pazhou Laboratory lizecheng19@gmail.com, {zeningzeng, auyqliang}@mail.scut.edu.cn, jingangyu@scut.edu.cn
Pseudocode Yes Algorithm 1 Complete Instances Mining (CIM) strategy
Open Source Code Yes Our implementation will be made available at https://github.com/Zecheng Li19/CIM.
Open Datasets Yes Following previous methods, we also evaluate our method on PASCAL VOC 2012 [Everingham et al., 2010] and MS COCO [Lin et al., 2014] datasets.
Dataset Splits Yes The VOC 2012 dataset includes 10,582 images for training and 1,449 images for validation, comprising 20 object categories. The COCO dataset comprises 115K training, 5K validation, and 20K testing images across 80 object categories.
Hardware Specification Yes Our method is implemented in Py Torch and experiments are conducted on an Nvidia RTX 3090.
Software Dependencies No The paper mentions 'Py Torch' but does not specify a version number or any other software dependencies with their versions.
Experiment Setup Yes We use COB [Maninis et al., 2018] method to generate proposals for all experiments and utilize Res Net50 [He et al., 2016] as the backbone. As we using m AP25 and m AP50 as evaluation metrics, we set classification τcls and integrity τiou thresholds to 0.25 and 0.5, respectively. The cascaded threshold τcas is set to 0.1. τnms, and pseed in Algorithm 1 are set to τcls, and 0.1, respectively. Containment threshold τcon is set to 0.85 following So S [Sui et al., 2021]. During training, we use the SGD optimization algorithm with an initial learning rate of 2.5 10 4 and a weight decay of 5 10 4. We adopt a step learning rate decay schema with a decay weight of 0.1 and set the mini-batch size to 4. The total number of training iterations is 4.5 104 for the VOC 2012 dataset and 24 104 iterations for the COCO dataset. For data augmentation, we apply five image scales {480, 576, 688, 864, 1200} with random horizontal flips for both training and testing.