SGAT4PASS: Spherical Geometry-Aware Transformer for PAnoramic Semantic Segmentation

Authors: Xuewei Li, Tao Wu, Zhongang Qi, Gaoang Wang, Ying Shan, Xi Li

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on Stanford2D3D Panoramic datasets show that SGAT4PASS significantly improves performance and robustness, with approximately a 2% increase in m Io U, and when small 3D disturbances occur in the data, the stability of our performance is improved by an order of magnitude.
Researcher Affiliation Collaboration 1College of Computer Science and Technology, Zhejiang University 2ARC Lab, Tencent PCG 3Zhejiang University-University of Illinois at Urbana-Champaign Institute, Zhejiang University 4Zhejiang Singapore Innovation and AI Joint Research Lab, Hangzhou
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code Yes Our code and supplementary material are available at https: //github.com/Tencent ARC/SGAT4PASS.
Open Datasets Yes We validate SGAT4PASS on Stanford2D3D Panoramic datasets [Armeni et al., 2017].
Dataset Splits Yes It has 1,413 panoramas, and 13 semantic classes are labeled, and has 3 official folds, fold 1 / 2 / 3. We follow the report style of previous work [Zhang et al., 2022a] [Zhang et al., 2022b].
Hardware Specification Yes Our experiments are conducted with a server with four A100 GPUs.
Software Dependencies No The paper mentions using 'Adam W' as an optimizer but does not specify version numbers for other software dependencies like Python, PyTorch, or CUDA.
Experiment Setup Yes We use Trans4PASS+ [Zhang et al., 2022b] as our baseline and set an initial learning rate of 8e-5, which is scheduled by the poly strategy with 0.9 power over 150 epochs. The optimizer is Adam W [Kingma and Ba, 2015] with epsilon 1e-8, weight decay 1e-4, and batch size is 4 on each GPU. Other settings and hyperparameters are set the same as Trans4PASS+ [Zhang et al., 2022b].