Evaluating Brush Movements for Chinese Calligraphy: A Computer Vision Based Approach

Authors: Pengfei Xu, Lei Wang, Ziyu Guan, Xia Zheng, Xiaojiang Chen, Zhanyong Tang, Dingyi Fang, Xiaoqing Gong, Zheng Wang

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

Reproducibility Variable Result LLM Response
Research Type Experimental We conducted extensive experiments and user studies to evaluate our approach. Experimental results show that our approach is highly accurate in identifying brush movements, yielding an average accuracy of 90%, and the generated score is within 3% of errors when compared to the one given by human experts.
Researcher Affiliation Academia 1 School of Information Science and Technology, Northwest University, China 2 Department of Cultural Heritage and Museology, Zhejiang University, China 3 Meta Lab, School of Computing and Communications, Lancaster University, U. K.
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code Yes Code and data are available at https://goo.gl/X1J2q R.
Open Datasets Yes Code and data are available at https://goo.gl/X1J2q R.
Dataset Splits No We randomly select 300 videos from the whole set as the training data by labeling the frame sequences of SBTP, writing a stroke and LBFP, and the remaining videos are used as the test data.This process is repeated 10 times.
Hardware Specification Yes Our approach runs on a workstation with an Intel E5 CPU, a Nvidia GTX 1080 GPU and 64GB of RAM.
Software Dependencies No The paper mentions software components like 'fast RCNN', 'TLD', 'DTW', 'MCNN-LSTM', 'CNN', but does not provide specific version numbers for any of these or other ancillary software dependencies.
Experiment Setup Yes The structure of convolution layers in MCNN-LSTM is design as mentioned above, batch size is 50 and learning rate is set as 0.001.