Semantic-Aware Generation of Multi-View Portrait Drawings

Authors: Biao Ma, Fei Gao, Chang Jiang, Nannan Wang, Gang Xu

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results show that SAGE achieves significantly superior or highly competitive performance, compared to existing 3Daware image synthesis methods.
Researcher Affiliation Academia 1 School of Computer Science and Technology, Hangzhou Dianzi University 2 Hangzhou Institute of Technology, Xidian University 3 ISN State Key Laboratory, Xidian University {aiartma, jc233, gxu}@hdu.edu.cn, {fgao, nnwang}@xidian.edu.cn
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code Yes The codes are available at https://github.com/Ai Art-HDU/SAGE.
Open Datasets Yes Pen-drawings. We conduct experiments on the APDrawing [Yi et al., 2019] dataset... Pencil-drawings. We apply our method to the three styles of pencil-drawings on the FS2K [Fan et al., 2022] dataset. ... Oil paintings. We randomly select 3,133 oil-paintings of humans from Wiki Art [Nichol, 2016]... Facial Photos. We use the Celeb AMask-HQ [Lee et al., 2020] dataset during training stage-I and for data augmentation.
Dataset Splits No The paper discusses training and testing datasets but does not explicitly provide details about a separate validation split, percentages, or sample counts for validation data.
Hardware Specification Yes We use a single Ge Force RTX3090 to train our model.
Software Dependencies No The paper mentions software components and optimizers (e.g., Adam optimizer, Ne RF) but does not provide specific version numbers for any libraries, frameworks, or other software dependencies.
Experiment Setup Yes In the pre-training phase of the model, we set λ1 = 0.1, λ2 = 1, and λ3 = 0.25. In the second phase of training, the parameters remain unchanged. We use the Adam optimizer and set β1 = 0 and β2 = 0.9. At 642 resolution, the learning rate of generator is 6 10 5, the learning rate of discriminators 2 10 4, and batch size 36. At 1282 resolution, the learning rate of generator is 5 10 5 and batch size 24. At 2562 resolution, the learning rate of generator is 3 10 5, the learning rate of discriminators 1 10 4, and batch size 24.