Benign Examples: Imperceptible Changes Can Enhance Image Translation Performance
Authors: Vignesh Srinivasan, Klaus-Robert Müller, Wojciech Samek, Shinichi Nakajima5842-5850
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Quantitative evaluation shows that our proposed method leads to a substantial increase in the accuracy to the target label on multiple state-of-the-art image classifiers, while qualitative user study proves that our method better represents the target domain, achieving better human preference scores. |
| Researcher Affiliation | Academia | Vignesh Srinivasan,1 Klaus-Robert M uller,2,3,4,5, Wojciech Samek,1,3, Shinichi Nakajima2,3,6, 1Fraunhofer HHI, 2TU Berlin, 3Berlin Big Data Center, 4Korea University, 5MPI for Informatics, 6RIKEN AIP |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper refers to publicly available pretrained models of Cycle GAN (Zhu et al. 2017) on Github, but it does not provide a link to its own source code for the methodology described in this paper. |
| Open Datasets | Yes | All the classifiers VGG16 (Simonyan and Zisserman 2014), Resnet18 (He et al. 2016), Inception V3 (Szegedy et al. 2016) and Resnet50 (Xie et al. 2017), Resnet101 (Zagoruyko and Komodakis 2016), Dense Net169 and Dense Net201 (Huang et al. 2017) have been trained on the Imagenet dataset (Deng et al. 2009) which includes a label for Zebras. The resulting Top-1 accuracy on the fake zebra images from Cycle GAN is shown in the second column in Table 1. ... we used the map dataset (Isola et al. 2017) consisting of satellite images and corresponding maps. |
| Dataset Splits | No | The paper mentions using |
| Hardware Specification | No | The paper does not provide specific details about the hardware used to run the experiments, such as GPU models or CPU specifications. |
| Software Dependencies | No | The paper mentions using |
| Experiment Setup | Yes | In both the cases, a step size of α = 0.01 and standard deviation of σ = 0.001 with a fixed number of iterations N = 50 was used for the Langevin dynamics. |