Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
High-Level Semantic Feature Matters Few-Shot Unsupervised Domain Adaptation
Authors: Lei Yu, Wanqi Yang, Shengqi Huang, Lei Wang, Ming Yang
AAAI 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on Domain Net show that the proposed method significantly outperforms SOTA methods in FS-UDA by a large margin (i.e., 10%). |
| Researcher Affiliation | Academia | 1School of Computer and Electronic Information, Nanjing Normal University, China 2School of Computing and Information Technology, University of Wollongong, Australia |
| Pseudocode | No | The paper describes methodological steps but does not include structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide an explicit statement about the release of source code or a link to a code repository for the described methodology. |
| Open Datasets | Yes | We conduct extensive experiments on a multi-domain benchmark dataset Domain Net to demonstrate the efficacy of our method. It was released in 2019 for the research of multi-source domain adaptation (Peng et al. 2019). |
| Dataset Splits | Yes | The dataset is split into 217, 43 and 48 categories for episodic training, model validation and testing new tasks, respectively. |
| Hardware Specification | No | The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | In cross-domain self-training module, we set the threshold 1.7 of similarity score to select the confidence samples in target domain. The margin m in Eq. (3) is empirically set to 1.5. ... The hyperparameters λsfa, λspa and λclm are set to 0.1, 0.05 and 0.01, by grid search, respectively. Also, we employ Adam optimizer with an initial learning rate of 10-4, and meanwhile reduce the learning rate by half every 280 episodes. |