LEARNING FACTORIZED REPRESENTATIONS FOR OPEN-SET DOMAIN ADAPTATION
Authors: Mahsa Baktashmotlagh, Masoud Faraki, Tom Drummond, Mathieu Salzmann
ICLR 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our experiments on standard benchmarks evidence that our approach outperforms the state of the art in open-set domain adaptation. |
| Researcher Affiliation | Academia | Mahsa Baktashmotlagh University of Queensland Masoud Faraki Monash University Tom Drummond* Monash University Mathieu Salzmann EPFL |
| Pseudocode | Yes | The pseudo-code of FRODA is provided in Algorithm 1. |
| Open Source Code | No | The paper does not contain an explicit statement about releasing source code or a link to a code repository. |
| Open Datasets | Yes | We evaluate our approach on the task of open-set visual domain adaptation using two benchmark datasets... namely Bing (B), Caltech256 (C), Image Net (I) and SUN (S), hence referred to as BCIS... on the Office dataset (Saenko et al., 2010). |
| Dataset Splits | Yes | We follow the unsupervised protocol of Tommasi & Tuytelaars (2014), which relies on 50 source samples per class and 30 target images per class, except when the target data is coming from SUN, in which case only 20 images per class are employed. |
| Hardware Specification | No | The paper states that "these runtimes were measured on the same computer" but does not provide any specific hardware details such as CPU model, GPU model, or memory. |
| Software Dependencies | No | The paper mentions using "De CAF7 features" but does not specify any software names with version numbers for libraries, frameworks, or programming languages used in the implementation. |
| Experiment Setup | Yes | For all our experiments, the hyperparameters of our approach were set as follows: α = 0.1, β = 0.01, λ1 = 0.001, λ2 = 0.001 and ε = 0.2. |