Style-Content Metric Learning for Multidomain Remote Sensing Object Recognition

Authors: Wenda Zhao, Ruikai Yang, Yu Liu, You He

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

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on four datasets show that our style-content metric learning achieves superior generalization performance against the state-of-the-art competitors.
Researcher Affiliation Academia 1Dalian University of Technology, Dalian, China 2Tsinghua University, Beijing, China
Pseudocode No The paper describes the framework and its components using text and equations, but does not provide a formal pseudocode or algorithm block.
Open Source Code Yes Code and model are available at: https://github.com/wdzhao123/TSCM.
Open Datasets Yes We conduct experiments using four remote sensing datasets: NWPU (Cheng, Zhou, and Han 2016), DOTA (Xia et al. 2018), HRRSD (Zhang et al. 2019) and DIOR (Li et al. 2020c).
Dataset Splits No The paper describes using training and testing datasets but does not provide explicit details for a separate validation split (e.g., percentages or sample counts).
Hardware Specification Yes Our model is implemented by Pytorch on a PC with a NVIDIA RTX 2080 Ti GPU.
Software Dependencies No The paper mentions 'Pytorch' and 'Adam' but does not specify version numbers for these or any other software dependencies.
Experiment Setup Yes We resize the image size to 256 x 256 pixels and set the batch size to 36. Adam (Kingma and Ba 2014) is used as the optimizer, and learning rate is 1.25e-4. We exponentially decay the learning rate of each parameter group by gamma set as 0.99 every epoch. Bias ε is set to 1e-6 in case that divisor and square root turn zero. The hyper-parameters are set as α = 0.1, β = 0.5 and γ = 0.5. The model is firstly trained for 48 epochs and then the last four fully-connected layers are further finetuned for 10 epochs to improve performance.