Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment

Authors: Yuanfan Zheng, Jinlin Wu, Wuyang Li, Zhen Chen

AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments demonstrate that our DPA outperforms state-of-the-art Uni DAOD and DAOD methods across various datasets and scenarios, including open, partial, and closed sets.
Researcher Affiliation Academia Yuanfan Zheng1,2*, Jinlin Wu1,2*, Wuyang Li3, Zhen Chen1 1CAIR, HKISI-CAS 2MAIS, Institute of Automation, Chinese Academy of Sciences 3The Chinese University of Hong Kong EMAIL, EMAIL, EMAIL, EMAIL
Pseudocode No The paper describes its methodology in prose and mathematical formulations but does not contain a clearly labeled pseudocode block or algorithm section.
Open Source Code Yes Code https://github.com/zyfone/DPA
Open Datasets Yes We evaluate our DPA framework on five datasets across three domain adaptation scenarios (open-set, partial-set, and closed-set): Foggy Cityscapes (Sakaridis, Dai, and Van Gool 2018), Cityscapes (Cordts et al. 2016), Pascal VOC (Everingham et al. 2010), Clipart1k (Inoue et al. 2018), and Watercolor (Inoue et al. 2018).
Dataset Splits No The paper states: "We conduct extensive experiments following the setting (Shi et al. 2022) for three benchmarks: open-set, partial-set, and closed-set." While it references a general experimental setting, it does not explicitly detail the training, validation, or test dataset splits (e.g., percentages, sample counts) for each dataset within the paper itself.
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments, such as GPU models, CPU types, or memory specifications.
Software Dependencies No The paper mentions optimizers like SGD and Adam but does not provide specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions).
Experiment Setup Yes The DPA model optimized training iterations are 100k, with an initial learning rate of 1e-3 and a subsequent decay of the learning rate to 1e-4 following 50k iterations. The LDPA is optimized using the SGD optimizer. The bound loss Lbound is optimized using the Adam optimizer with a learning rate set to 0.1. The hyperparameter of α is 0 for the initial epoch and 0.1 thereafter.