Face Sketch Synthesis From Coarse to Fine

Authors: Mingjin Zhang, Nannan Wang, Yunsong Li, Ruxin Wang, Xinbo Gao

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on public face sketch databases illustrate that our proposed framework outperforms the state-of-the-art methods in both quantitive and visual comparisons.
Researcher Affiliation Collaboration 1 State Key Laboratory of Integrated Services Networks, School of Telecommunications, Xidian University, Xi an 710071, China 2 Yunnan Union Vision Innovations Technology Company Limited, Kunming 650000, China 3 School of Electronic Engineering, Xidian University, Xi an 710071, China
Pseudocode Yes Algorithm 1 Realization procedure
Open Source Code No The paper does not provide any explicit statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes We conduct experiments on the Chinese University of Hong Kong (CUHK) face sketch database (CUFS) (Wang and Tang 2009), which consists of 606 face sketchphoto pairs.
Dataset Splits No The paper describes training and test splits for the datasets but does not explicitly mention a separate validation split or how it was used.
Hardware Specification Yes The coarse stage is conducted using Torch on Ubuntu 14.04 system with 12G NVIDIA Titan X GPU, whereas the fine stage is tested using Matlab on Window 7 System with i7-4790 3.6G CPU.
Software Dependencies No The paper mentions 'Torch' and 'Matlab' as software used, but does not provide specific version numbers for these software dependencies, only for the operating systems (Ubuntu 14.04, Window 7).
Experiment Setup Yes In the proposed approach, we set six parameters. The number of candidates K is 10, the patch size is 11, the overlap size is 7, the search region is 5, and the balance parameters α and β are 100 and 1, respectively.