SFC: Shared Feature Calibration in Weakly Supervised Semantic Segmentation
Authors: Xinqiao Zhao, Feilong Tang, Xiaoyang Wang, Jimin Xiao
AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results show that our SFC significantly improves CAM boundaries and achieves new state-of-the-art performances. |
| Researcher Affiliation | Collaboration | 1Xi an Jiaotong-Liverpool University 2University of Liverpool 3Metavisioncn xqz@liverpool.ac.uk, Feilong.Tang19@xjtlu.edu.cn, wangxy@liverpool.ac.uk, Jimin.Xiao@xjtlu.edu.cn |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks that are clearly labeled as such. |
| Open Source Code | Yes | The project is available at https://github.com/Barrett-python/SFC. |
| Open Datasets | Yes | Experiments are conducted on two benchmarks: PASCAL VOC 2012 (Everingham et al. 2010) with 21 classes and MS COCO 2014 (Lin et al. 2014) with 81 classes. For PASCAL VOC 2012, following (Wang et al. 2020; Lee, Kim, and Yoon 2021; Chen et al. 2022a; Li et al. 2022), we use the augmented SBD set (Hariharan et al. 2011) with 10,582 annotated images. |
| Dataset Splits | Yes | Table 2 reports the m Io U scores of our method and recent WSSS methods on the validation and test sets of PASCAL VOC 2012. |
| Hardware Specification | No | The paper does not provide specific hardware details such as exact GPU/CPU models or processor types used for running its experiments. |
| Software Dependencies | No | The paper mentions using ImageNet pretrained Res Net50, DeepLab V2, but does not provide specific version numbers for software dependencies like Python, PyTorch, or other libraries. |
| Experiment Setup | Yes | Random cropping size 512 512 is adopted for training data augmentation. The NIBR is set to 4. Mfinal from our method is further post-processed by Dence CRF (Kr ahenb uhl and Koltun 2011) and IRN (Ahn, Cho, and Kwak 2019) to generate the final pseudo labels, which are used to train the segmentation model: Res Net101-based Deep Lab V2 (Ahn, Cho, and Kwak 2019; Chen et al. 2022a). |