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].
ProMEA: Prompt-driven Expansion and Alignment for Single Domain Generalization
Authors: Yunyun Wang, Yi Guo, Xiaodong Liu, Songcan Chen
IJCAI 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirically results over single-DG benchmarks demonstrate the superiority of our proposal. |
| Researcher Affiliation | Academia | Yunyun Wang1 , Yi Guo1 , Xiaodong Liu1 and Songcan Chen2,3 1School of Computer Science, University of Posts and Telecommunications 2 College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics 3 MIIT Key Laboratory of Pattern Analysis and Machine Intelligence, Nanjing University of Aeronautics and Astronautics EMAIL, EMAIL |
| Pseudocode | Yes | The learning process is summarized in Algorithm 1. Algorithm 1 Training of Pro MEA Input: Source domain Ds, domain words Wd = {d1, . . . , d K}, prompt templates Prompts, and CLIP text encoder Etxt( ) Output: Learned model weights θ 1: for e Max Epoch do 2: ztxt kt, zstyle k Prompts, Wd initialize text prompts 3: Calculate Lds according to Eq. 3 4: xaug ik F 1(A(ztxt kt ), P(xi)) synthesize samples 5: for domain word dk Wd do 6: ωik similarity(xi, zstyle k ) 7: end for 8: Amix i 1 K PK k=1 ωik A zstyle k 9: xmix i F 1((Amix i ), P(xi)) 10: Calculate Laug cls , Lmix cls according to Eq. 8,12 11: if warm up then 12: update θ to minimize Lcls + αLaug cls + λLds 13: else 14: update θ to minimize L according to Eq. 13 15: end if 16: end for |
| Open Source Code | No | The paper does not contain any explicit statement about releasing source code for the methodology described, nor does it provide a link to a code repository. |
| Open Datasets | Yes | 4.1 Datasets PACS [Li et al., 2017a] is composed of four domains (Photo, Art, Cartoon, Sketch) and 7 classes (dog, elephant, giraffe, guitar, horse and person), and there is large distribution discrepancy between different domains. VLCS [Fang et al., 2013] consists of 10,729 images across 5 classes (bird, car, chair, dog, and person), sourced from 4 real-world datasets (VOC2007, Label Me, Caltech, SUN09). The scenes captured vary from urban to rural. Office Home [Venkateswara et al., 2017] is a challenging dataset. It contains 30,475 images belonging to 65 classes, originating from four different domains (Art, Clipart, Product, Real), where the domain shift mainly stems from differences in image styles and viewpoints. |
| Dataset Splits | Yes | For VLCS, we randomly split each domain into 90% training and 10% validation subsets. For PACS and Office Home, we follow the data splits from Shu et al. [Shu et al., 2021]. The model is trained on a single source domain and evaluated on the remaining domains. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU models, CPU models, or memory amounts used for running experiments. It mentions using ResNet-18 and ResNet-50 as backbones but not the hardware on which these were run. |
| Software Dependencies | No | The paper mentions using a 'pre-trained CLIP model' and 'SGD optimizer' but does not specify version numbers for any software libraries, programming languages (like Python), or other dependencies. |
| Experiment Setup | Yes | We use SGD optimizer for training, set the batch size to 16, the weight decay to 5e-4, and train for 75 epochs. The initial learning rate is 1e-3, and the warm-up period consists of 25 epochs. For Office Home, we choose Res Net-50 as the backbone and set the weight decay to 5e-5, training for 50 epochs. We train with SGD optimizer and set the batch size to 16. The initial learning rate is 1e-3 and decays by 0.1 at 80% of the total epochs. The warm-up period lasts for 10 epochs. In the experiments, we initially set α = 0.3, β = 0.5, λ = 0.01 and the warm-up epoch at 10. |